Llama 2: Open Foundation and Fine-Tuned Chat Models

2023/07/18 by Hugo Touvron, Louis Martin, Touvron, Hugo +133 · 1306 citations
Computer Science · #Topic Modeling #Natural Language Processing Techniques #Speech and dialogue systems

paper · pdf · doi:10.48550/arxiv.2307.09288

Abstract

In this work, we develop and release Llama 2, a collection of pretrained and fine-tuned large language models (LLMs) ranging in scale from 7 billion to 70 billion parameters. Our fine-tuned LLMs, called Llama 2-Chat, are optimized for dialogue use cases. Our models outperform open-source chat models on most benchmarks we tested, and based on our human evaluations for helpfulness and safety, may be a suitable substitute for closed-source models. We provide a detailed description of our approach to fine-tuning and safety improvements of Llama 2-Chat in order to enable the community to build on our work and contribute to the responsible development of LLMs.

Cited by

• DeCoRAG: Cognitive Decoupling and Semantic-Aware Cropping for Complex Document Understanding
• PIVOT: Efficient Query-Group Indexing for Token-Level Sparse Attention
• Instruction-Following Evaluation of Large Vision-Language Models
• Scoring, Reasoning, and Selecting the Best! Ensembling Large Language Models via a Peer-Review Process
• Reservoir Computing inspired Matrix Multiplication-free Language Model
• Rethinking Fine-Tuning: Unlocking Hidden Capabilities in Vision-Language Models
• Merge before Forget: A Single LoRA Continual Learning via Continual Merging
• Towards Understanding Steering Strength
• Eliciting Behaviors in Multi-Turn Conversations
• HiSciBench: A Hierarchical Multi-disciplinary Benchmark for Scientific Intelligence from Reading to Discovery
• EmoCtrl: Controllable Emotional Image Content Generation
• Agent2World: Learning to Generate Symbolic World Models via Adaptive Multi-Agent Feedback
• Learning from Negative Examples: Why Warning-Framed Training Data Teaches What It Warns Against
• Calibrating LLM Judges: Linear Probes for Fast and Reliable Uncertainty Estimation
• On the Existence and Behavior of Secondary Attention Sinks
• Open-Source Multimodal Moxin Models with Moxin-VLM and Moxin-VLA
• Latent-LoRA: Compact Latent-Space Adapters with Gradient-Free Routing for Continual Learning
• TriShieldRAG: A Three-Ring Defense-in-Depth Framework Against Knowledge Corruption in Retrieval-Augmented Generation
• Scale Weight Decay and Train Better
• DeepFaith: Evidence-Grounded LLMs for Faithful Incident Reporting in Multi-Stage APT Defense
• D-Score: A Spectral Hidden-State Signal for Hallucination Detection in Large Language Models
• Every Time I Hire a Linguist, Inference Costs Go Down: On Linguistic Rules as Effective Prompt Compressors
• MixQuant: Adaptive Mixed-Precision Quantization for Large Language Models
• Unifying Learning Dynamics and Generalization in Transformers Scaling Law
• Evaluating large language models for diagnostic reasoning from unstructured clinical narratives in epilepsy
• When Sinks Help or Hurt: Unified Framework for Attention Sink in Large Vision-Language Models
• SafeCRS: Personalized Safety Alignment for LLM-Based Conversational Recommender Systems
• SkipOPU: An FPGA-based Overlay Processor for Large Language Models with Dynamically Allocated Computation
• RM-Distiller: Exploiting Generative LLM for Reward Model Distillation
• Compressing LLMs with MoP: Mixture of Pruners
• Responsible Intelligence in Practice: A Fairness Audit of Open Large Language Models for Library Reference Services
• PixelGen: Improving Pixel Diffusion with Perceptual Supervision
• Training-Free Adaptation of New-Generation LLMs using Legacy Clinical Models
• Explainability Methods for Hardware Trojan Detection: A Systematic Comparison
• PINA: Prompt Injection Attack Against Navigation Agents
• DPAR: Dynamic Patchification for Efficient Autoregressive Visual Generation
• Fast Inference of Visual Autoregressive Model with Adjacency-Adaptive Dynamical Draft Trees
• A Comedy of Estimators: On KL Regularization in RL Training of LLMs
• Rethinking Output Alignment For 1-bit Post-Training Quantization of Large Language Models
• Gamayun's Path to Multilingual Mastery: Cost-Efficient Training of a 1.5B-Parameter LLM
• Parallel Token Prediction for Language Models
• Casting a SPELL: Sentence Pairing Exploration for LLM Limitation-breaking
• ORCA: Object Recognition and Comprehension for Archiving Marine Species
• SPOT!: Map-Guided LLM Agent for Unsupervised Multi-CCTV Dynamic Object Tracking
• ALIVE: An Avatar-Lecture Interactive Video Engine with Content-Aware Retrieval for Real-Time Interaction
• Generative Digital Twins: Vision-Language Simulation Models for Executable Industrial Systems
• Beyond Language Boundaries: Uncovering Programming Language Families for Code Language Models
• Activations as Features: Probing LLMs for Generalizable Essay Scoring Representations
• HATS: High-Accuracy Triple-Set Watermarking for Large Language Models
• ReasonCD: A Multimodal Reasoning Large Model for Implicit Change-of-Interest Semantic Mining
• The Mental World of Large Language Models in Recommendation: A Benchmark on Association, Personalization, and Knowledgeability
• MAGIC: Achieving Superior Model Merging via Magnitude Calibration
• Learning Hierarchical Procedural Memory for LLM Agents through Bayesian Selection and Contrastive Refinement
• Learning-Based Automated Adversarial Red-Teaming for Robustness Evaluation of Large Language Models
• MoE Pathfinder: Trajectory-driven Expert Pruning
• DACE For Railway Acronym Disambiguation
• Towards Benchmarking Privacy Vulnerabilities in Selective Forgetting with Large Language Models
• When the Gold Standard Isn't Necessarily Standard: Challenges of Evaluating the Translation of User-Generated Content
• Towards Deeper Emotional Reflection: Crafting Affective Image Filters with Generative Priors
• Incorporating Error Level Noise Embedding for Improving LLM-Assisted Robustness in Persian Speech Recognition
• Physics of Language Models: Part 4.1, Architecture Design and the Magic of Canon Layers
• UniRel: Relation-Centric Knowledge Graph Question Answering with RL-Tuned LLM Reasoning
• 4D-RGPT: Toward Region-level 4D Understanding via Perceptual Distillation
• Posterior Behavioral Cloning: Pretraining BC Policies for Efficient RL Finetuning
• AdaSearch: Balancing Parametric Knowledge and Search in Large Language Models via Reinforcement Learning
• Toward Systematic Counterfactual Fairness Evaluation of Large Language Models: The CAFFE Framework
• Efficient CPU-GPU Collaborative Inference for MoE-based LLMs on Memory-Limited Systems
• Collaborative Edge-to-Server Inference for Vision-Language Models
• Design and Evaluation of Cost-Aware PoQ for Decentralized LLM Inference
• Fast Collaborative Inference via Distributed Speculative Decoding
• LoPA: Scaling dLLM Inference via Lookahead Parallel Decoding
• Mitigating Hallucinations in Healthcare LLMs with Granular Fact-Checking and Domain-Specific Adaptation
• Vision-Language-Action Models for Autonomous Driving: Past, Present, and Future
• Spoken DialogSum: An Emotion-Rich Conversational Dataset for Spoken Dialogue Summarization
• LLMQ: Efficient Lower-Precision Pretraining for Consumer GPUs
• Beyond Fast and Slow: Cognitive-Inspired Elastic Reasoning for Large Language Models
• An Exploratory Study of Bayesian Prompt Optimization for Test-Driven Code Generation with Large Language Models
• PADE: A Predictor-Free Sparse Attention Accelerator via Unified Execution and Stage Fusion
• TEMP: A Memory Efficient Physical-aware Tensor Partition-Mapping Framework on Wafer-scale Chips
• Georeferencing complex relative locality descriptions with large language models
• Neurosymbolic Inference On Foundation Models For Remote Sensing Text-to-image Retrieval With Complex Queries
• OpenDataArena: A Fair and Open Arena for Benchmarking Post-Training Dataset Value
• Polypersona: Persona-Grounded LLM for Synthetic Survey Responses
• Autonomous Construction-Site Safety Inspection Using Mobile Robots: A Multilayer VLM-LLM Pipeline
• Counting Clues: A Lightweight Probabilistic Baseline Can Match an LLM
• Understanding Structured Financial Data with LLMs: A Case Study on Fraud Detection
• SliceMoE: Bit-Sliced Expert Caching under Miss-Rate Constraints for Efficient MoE Inference
• FlashFuser: Expanding the Scale of Kernel Fusion for Compute-Intensive Operators via Inter-Core Connection
• CTIGuardian: A Few-Shot Framework for Mitigating Privacy Leakage in Fine-Tuned LLMs
• SkipCat: Rank-Maximized Low-Rank Compression of Large Language Models via Shared Projection and Block Skipping
• Informing Acquisition Functions via Foundation Models for Molecular Discovery
• Curió-Edu 7B: Examining Data Selection Impacts in LLM Continued Pretraining
• Leader-driven or Leaderless: How Participation Structure Sustains Engagement and Shapes Narratives in Online Hate Communities
• WATOS: Efficient LLM Training Strategies and Architecture Co-exploration for Wafer-scale Chip
• Rethinking Label Consistency of In-Context Learning: An Implicit Transductive Label Propagation Perspective
• Multi-Intent Spoken Language Understanding: Methods, Trends, and Challenges
• An Anatomy of Vision-Language-Action Models: From Modules to Milestones and Challenges
• qa-FLoRA: Data-free query-adaptive Fusion of LoRAs for LLMs
• VGent: Visual Grounding via Modular Design for Disentangling Reasoning and Prediction
• MeViS: A Multi-Modal Dataset for Referring Motion Expression Video Segmentation
• Stronger Normalization-Free Transformers
• Unlocking the Address Book: Dissecting the Sparse Semantic Structure of LLM Key-Value Caches via Sparse Autoencoders
• Openpi Comet: Competition Solution For 2025 BEHAVIOR Challenge
• Design Space Exploration of DMA based Finer-Grain Compute Communication Overlap
• XDoGE: Multilingual Data Reweighting to Enhance Language Inclusivity in LLMs
• GLaD: Geometric Latent Distillation for Vision-Language-Action Models
• Investigate the Low-level Visual Perception in Vision-Language based Image Quality Assessment
• Building Reasonable Inference for Vision-Language Models in Blind Image Quality Assessment
• Chasing Shadows: Pitfalls in LLM Security Research
• WarmServe: Enabling One-for-Many GPU Prewarming for Multi-LLM Serving
• MindShift: Analyzing Language Models' Reactions to Psychological Prompts
• Detecting Hallucinations in Graph Retrieval-Augmented Generation via Attention Patterns and Semantic Alignment
• Fed-SE: Federated Self-Evolution for Privacy-Constrained Multi-Environment LLM Agents
• Automatic Essay Scoring and Feedback Generation in Basque Language Learning
• Photo3D: Advancing Photorealistic 3D Generation through Structure-Aligned Detail Enhancement
• Attention is All You Need to Defend Against Indirect Prompt Injection Attacks in LLMs
• MolSculpt: Sculpting 3D Molecular Geometries from Chemical Syntax
• DermETAS-SNA LLM: A Dermatology Focused Evolutionary Transformer Architecture Search with StackNet Augmented LLM Assistant
• CVP: Central-Peripheral Vision-Inspired Multimodal Model for Spatial Reasoning
• ValuePilot: A Two-Phase Framework for Value-Driven Decision-Making
• TROJail: Trajectory-Level Optimization for Multi-Turn Large Language Model Jailbreaks with Process Rewards
• Training-Free Vector Quantization via Gaussian VAEs
• Persian-Phi: Efficient Cross-Lingual Adaptation of Compact LLMs via Curriculum Learning
• LUNE: Efficient LLM Unlearning via LoRA Fine-Tuning with Negative Examples
• Materium: An Autoregressive Approach for Material Generation
• See Once, Then Act: Vision-Language-Action Model with Task Learning from One-Shot Video Demonstrations
• ThinkTrap: Denial-of-Service Attacks against Black-box LLM Services via Infinite Thinking
• FOAM: Blocked State Folding for Memory-Efficient LLM Training
• Leveraging KV Similarity for Online Structured Pruning in LLMs
• Dual Refinement Cycle Learning: Unsupervised Text Classification of Mamba and Community Detection on Text Attributed Graph
• Large Language Model-Based Generation of Discharge Summaries
• Optimizing LLMs Using Quantization for Mobile Execution
• BitStopper: An Efficient Transformer Attention Accelerator via Stage-fusion and Early Termination
• Efficient Text Classification with Conformal In-Context Learning
• WAM-Flow: Parallel Coarse-to-Fine Motion Planning via Discrete Flow Matching for Autonomous Driving
• KQ-SVD: Compressing the KV Cache with Provable Guarantees on Attention Fidelity
• Optimizing Medical Question-Answering Systems: A Comparative Study of Fine-Tuned and Zero-Shot Large Language Models with RAG Framework
• Scaling and Transferability of Annealing Strategies in Large Language Model Training
• HiMoE-VLA: Hierarchical Mixture-of-Experts for Generalist Vision-Language-Action Policies
• The Road of Adaptive AI for Precision in Cybersecurity
• HiPPO: Exploring A Novel Hierarchical Pronunciation Assessment Approach for Spoken Languages
• GeoPE:A Unified Geometric Positional Embedding for Structured Tensors
• STELLA: Guiding Large Language Models for Time Series Forecasting with Semantic Abstractions
• Qwen3.5-Omni Technical Report
• Group Selection as a Safeguard Against AI Substitution
• SignRoundV2: Closing the Performance Gap in Extremely Low-Bit Post-Training Quantization for LLMs
• LeMat-GenBench: A Unified Evaluation Framework for Crystal Generative Models
• MSME: A Multi-Stage Multi-Expert Framework for Zero-Shot Stance Detection
• Vision-Language-Action Models for Selective Robotic Disassembly: A Case Study on Critical Component Extraction from Desktops
• Towards better dense rewards in Reinforcement Learning Applications
• D-STEER - Preference Alignment Techniques Learn to Behave, not to Believe -- Beneath the Surface, DPO as Steering Vector Perturbation in Activation Space
• Teaching Old Tokenizers New Words: Efficient Tokenizer Adaptation for Pre-trained Models
• Enhancing Instruction-Following Capabilities in Seq2Seq Models: DoLA Adaptations for T5
• LSRS: Latent Scale Rejection Sampling for Visual Autoregressive Modeling
• PosA-VLA: Enhancing Action Generation via Pose-Conditioned Anchor Attention
• Context-Aware Hierarchical Learning: A Two-Step Paradigm towards Safer LLMs
• KVNAND: Efficient On-Device Large Language Model Inference Using DRAM-Free In-Flash Computing
• Exploring the Potential and Limitations of Large Language Models for Novice Program Fault Localization
• Large Language Models as Generalist Policies for Network Optimization
• UniQL: Unified Quantization and Low-rank Compression for Adaptive Edge LLMs
• MarkTune: Improving the Quality-Detectability Trade-off in Open-Weight LLM Watermarking
• GeoZero: Incentivizing Reasoning from Scratch on Geospatial Scenes
• Fairy2i: Training Complex LLMs from Real LLMs with All Parameters in \± 1, ± i\
• Cross-Lingual Prompt Steerability: Towards Accurate and Robust LLM Behavior across Languages
• ESACT: An End-to-End Sparse Accelerator for Compute-Intensive Transformers via Local Similarity
• Masking Matters: Unlocking the Spatial Reasoning Capabilities of LLMs for 3D Scene-Language Understanding
• VACoT: Rethinking Visual Data Augmentation with VLMs
• Parameter-Efficient Subspace Optimization for LLM Fine-Tuning
• Four Over Six: More Accurate NVFP4 Quantization with Adaptive Block Scaling
• Ensemble Privacy Defense for Knowledge-Intensive LLMs against Membership Inference Attacks
• Script: Graph-Structured and Query-Conditioned Semantic Token Pruning for Multimodal Large Language Models
• Med-VCD: Mitigating Hallucination for Medical Large Vision Language Models through Visual Contrastive Decoding
• Edge Deployment of Small Language Models, a comprehensive comparison of CPU, GPU and NPU backends
• DiG-Flow: Discrepancy-Guided Flow Matching for Robust VLA Models
• RoMe: Row Granularity Access Memory System for Large Language Models
• Securing Large Language Models (LLMs) from Prompt Injection Attacks
• KVReviver: Reversible KV Cache Compression with Sketch-Based Token Reconstruction
• Stay Unique, Stay Efficient: Preserving Model Personality in Multi-Task Merging
• Table as a Modality for Large Language Models
• AFRAgent : An Adaptive Feature Renormalization Based High Resolution Aware GUI agent
• Elastic Mixture of Rank-Wise Experts for Knowledge Reuse in Federated Fine-Tuning
• Hybrid-DMKG: A Hybrid Reasoning Framework over Dynamic Multimodal Knowledge Graphs for Multimodal Multihop QA with Knowledge Editing
• Clinical-R1: Empowering Large Language Models for Faithful and Comprehensive Reasoning with Clinical Objective Relative Policy Optimization
• WaterSearch: A Quality-Aware Search-based Watermarking Framework for Large Language Models
• UMM-RM: An Upcycle-and-Merge MoE Reward Model for Mitigating Reward Hacking
• SpeContext: Enabling Efficient Long-context Reasoning with Speculative Context Sparsity in LLMs
• Optimizing LVLMs with On-Policy Data for Effective Hallucination Mitigation
• Financial Text Classification Based On rLoRA Finetuning On Qwen3-8B model
• ChartPoint: Guiding MLLMs with Grounding Reflection for Chart Reasoning
• DialBench: Towards Accurate Reading Recognition of Pointer Meter using Large Foundation Models
• Video-R2: Reinforcing Consistent and Grounded Reasoning in Multimodal Language Models
• TWEO: Transformers Without Extreme Outliers Enables FP8 Training And Quantization For Dummies
• SpaceMind: Camera-Guided Modality Fusion for Spatial Reasoning in Vision-Language Models
• Efficient Asynchronous Federated Evaluation with Strategy Similarity Awareness for Intent-Based Networking in Industrial Internet of Things
• Platinum: Path-Adaptable LUT-Based Accelerator Tailored for Low-Bit Weight Matrix Multiplication
• The Immutable Tensor Architecture: A Pure Dataflow Approach for Secure, Energy-Efficient AI Inference
• JarvisEvo: Towards a Self-Evolving Photo Editing Agent with Synergistic Editor-Evaluator Optimization
• Every Token Counts: Generalizing 16M Ultra-Long Context in Large Language Models
• UniGeoSeg: Towards Unified Open-World Segmentation for Geospatial Scenes
• Closed-Loop Transformers: Autoregressive Modeling as Iterative Latent Equilibrium
• Invisible Hands: Gray-Box Bit Flip Attack for Steering LLMs Without Knowledge of Gradients, Data, and Weights
• CacheTrap: Injecting Trojans in LLMs without Leaving any Traces in Inputs or Weights
• Generative models for crystalline materials
• Enhancing Trustworthiness with Mixed Precision: Benchmarks, Opportunities, and Challenges
• DSD: A Distributed Speculative Decoding Solution for Edge-Cloud Agile Large Model Serving
• Flowing Backwards: Improving Normalizing Flows via Reverse Representation Alignment
• SingleQuant: Efficient Quantization of Large Language Models in a Single Pass
• Swarms of Large Language Model Agents for Protein Sequence Design with Experimental Validation
• Co-Evolving Agents: Learning from Failures as Hard Negatives
• Attention-Guided Patch-Wise Sparse Adversarial Attacks on Vision-Language-Action Models
• E₀: Enhancing Generalization and Fine-Grained Control in VLA Models via Continuized Discrete Diffusion
• Bootstrapping LLMs via Preference-Based Policy Optimization
• ICPO: Intrinsic Confidence-Driven Group Relative Preference Optimization for Efficient Reinforcement Learning
• DOPD: A Dynamic PD-Disaggregation Architecture for Maximizing Goodput in LLM Inference Serving
• Graph-O1 : Monte Carlo Tree Search with Reinforcement Learning for Text-Attributed Graph Reasoning
• LAPA: Log-Domain Prediction-Driven Dynamic Sparsity Accelerator for Transformer Model
• Generating Querying Code from Text for Multi-Modal Electronic Health Record
• DRAFT-RL: Multi-Agent Chain-of-Draft Reasoning for Reinforcement Learning-Enhanced LLMs
• MapReduce LoRA: Advancing the Pareto Front in Multi-Preference Optimization for Generative Models
• Copyright Detection in Large Language Models: An Ethical Approach to Generative AI Development
• The Consistency Critic: Correcting Inconsistencies in Generated Images via Reference-Guided Attentive Alignment
• DesignPref: Capturing Personal Preferences in Visual Design Generation
• Large Language Models' Complicit Responses to Illicit Instructions across Socio-Legal Contexts
• Scaling LLM Speculative Decoding: Non-Autoregressive Forecasting in Large-Batch Scenarios
• REFLEX: Self-Refining Explainable Fact-Checking via Disentangling Truth into Style and Substance
• Vision-Language Models for Automated 3D PET/CT Report Generation
• Cross-Contrastive Clustering for Multimodal Attributed Graphs with Dual Graph Filtering
• Boosting Reasoning in Large Multimodal Models via Activation Replay
• EfficientXpert: Efficient Domain Adaptation for Large Language Models via Propagation-Aware Pruning
• It Hears, It Sees too: Multi-Modal LLM for Depression Detection By Integrating Visual Understanding into Audio Language Models
• A Systematic Analysis of Large Language Models with RAG-enabled Dynamic Prompting for Medical Error Detection and Correction
• The Curious Case of Analogies: Investigating Analogical Reasoning in Large Language Models
• DeCo: Frequency-Decoupled Pixel Diffusion for End-to-End Image Generation
• LAST: LeArning to Think in Space and Time for Generalist Vision-Language Models
• ABM-LoRA: Activation Boundary Matching for Fast Convergence in Low-Rank Adaptation
• FastForward Pruning: Efficient LLM Pruning via Single-Step Reinforcement Learning
• AVA-VLA: Improving Vision-Language-Action models with Active Visual Attention
• BackdoorVLM: A Benchmark for Backdoor Attacks on Vision-Language Models
• Nemotron-Flash: Towards Latency-Optimal Hybrid Small Language Models
• CoreEval: Automatically Building Contamination-Resilient Datasets with Real-World Knowledge toward Reliable LLM Evaluation
• Low-Rank GEMM: Efficient Matrix Multiplication via Low-Rank Approximation with FP8 Acceleration
• Defending Large Language Models Against Jailbreak Exploits with Responsible AI Considerations
• An Invariant Latent Space Perspective on Language Model Inversion
• TASO: Jailbreak LLMs via Alternative Template and Suffix Optimization
• Foundations of Artificial Intelligence Frameworks: Notion and Limits of AGI
• ChineseVideoBench: Benchmarking Multi-modal Large Models for Chinese Video Question Answering
• ARIAL: An Agentic Framework for Document VQA with Precise Answer Localization
• Vector Arithmetic in Concept and Token Subspaces
• Consolidating Diffusion-Generated Video Detection with Unified Multimodal Forgery Learning
• MGA-VQA: Secure and Interpretable Graph-Augmented Visual Question Answering with Memory-Guided Protection Against Unauthorized Knowledge Use
• Generative Adversarial Post-Training Mitigates Reward Hacking in Live Human-AI Music Interaction
• Show Me: Unifying Instructional Image and Video Generation with Diffusion Models
• PoETa v2: Toward More Robust Evaluation of Large Language Models in Portuguese
• Layer-Wise High-Impact Parameter Ratio Optimization in Post-Training Quantization for Large Language Models
• PersonaAgent with GraphRAG: Community-Aware Knowledge Graphs for Personalized LLM
• SAVeD: Semantic Aware Version Discovery
• Estonian WinoGrande Dataset: Comparative Analysis of LLM Performance on Human and Machine Translation
• Intervene-All-Paths: Unified Mitigation of LVLM Hallucinations across Alignment Formats
• LangMark: A Multilingual Dataset for Automatic Post-Editing
• VLA-4D: Embedding 4D Awareness into Vision-Language-Action Models for SpatioTemporally Coherent Robotic Manipulation
• Bridging the Semantic-Action Gap in Visual Token Pruning for Efficient VLA Inference
• Adaptive Layer-Wise Transformations for Post-Training Quantization of Large Language Models
• METIS: Multi-Source Egocentric Training for Integrated Dexterous Vision-Language-Action Model
• Nemotron Elastic: Towards Efficient Many-in-One Reasoning LLMs
• When Structure Doesn't Help: LLMs Do Not Read Text-Attributed Graphs as Effectively as We Expected
• You Only Forward Once: An Efficient Compositional Judging Paradigm
• TimeViper: A Hybrid Mamba-Transformer Vision-Language Model for Efficient Long Video Understanding
• AICC: Parse HTML Finer, Make Models Better -- A 7.3T AI-Ready Corpus Built by a Model-Based HTML Parser
• ELPO: Ensemble Learning Based Prompt Optimization for Large Language Models
• "To Survive, I Must Defect": Jailbreaking LLMs via the Game-Theory Scenarios
• AMS-KV: Adaptive KV Caching in Multi-Scale Visual Autoregressive Transformers
• Global Resolution: Optimal Multi-Draft Speculative Sampling via Convex Minimization
• Explore How to Inject Beneficial Noise in MLLMs
• Automatic Pruning Discovery for Large Language Models
• As If We've Met Before: LLMs Exhibit Certainty in Recognizing Seen Files
• GRPO-RM: Fine-Tuning Representation Models via GRPO-Driven Reinforcement Learning
• OEMA: Ontology-Enhanced Multi-Agent Collaboration Framework for Zero-Shot Clinical Named Entity Recognition
• Multi-Aspect Cross-modal Quantization for Generative Recommendation
• Strategic Innovation Management in the Age of Large Language Models Market Intelligence, Adaptive R&D, and Ethical Governance
• Steganographic Backdoor Attacks in NLP: Ultra-Low Poisoning and Defense Evasion
• Orion: A Unified Visual Agent for Multimodal Perception, Advanced Visual Reasoning and Execution
• Beyond Surface-Level Similarity: Hierarchical Contamination Detection for Synthetic Training Data in Foundation Models
• Insight-A: Attribution-aware for Multimodal Misinformation Detection
• T-SAR: A Full-Stack Co-design for CPU-Only Ternary LLM Inference via In-Place SIMD ALU Reorganization
• Building Egocentric Procedural AI Assistant: Methods, Benchmarks, and Challenges
• Souper-Model: How Simple Arithmetic Unlocks State-of-the-Art LLM Performance
• Evaluating Large Language Models for Diacritic Restoration in Romanian Texts: A Comparative Study
• OTARo: Once Tuning for All Precisions toward Robust On-Device LLMs
• MACKO: Sparse Matrix-Vector Multiplication for Low Sparsity
• PixelVLA: Advancing Pixel-level Understanding in Vision-Language-Action Model
• EVLP:Learning Unified Embodied Vision-Language Planner with Reinforced Supervised Fine-Tuning
• Medical Knowledge Intervention Prompt Tuning for Medical Image Classification
• CSV-Decode: Certifiable Sub-Vocabulary Decoding for Efficient Large Language Model Inference
• Mitigating Length Bias in RLHF through a Causal Lens
• Seg-VAR: Image Segmentation with Visual Autoregressive Modeling
• GenSIaC: Toward Security-Aware Infrastructure-as-Code Generation with Large Language Models
• Rethinking Deep Alignment Through The Lens Of Incomplete Learning
• OAD-Promoter: Enhancing Zero-shot VQA using Large Language Models with Object Attribute Description
• PRISM of Opinions: A Persona-Reasoned Multimodal Framework for User-centric Conversational Stance Detection
• GateRA: Token-Aware Modulation for Parameter-Efficient Fine-Tuning
• Align³GR: Unified Multi-Level Alignment for LLM-based Generative Recommendation
• Defending Unauthorized Model Merging via Dual-Stage Weight Protection
• Aligning Machiavellian Agents: Behavior Steering via Test-Time Policy Shaping
• PAS : Prelim Attention Score for Detecting Object Hallucinations in Large Vision--Language Models
• CURENet: Combining Unified Representations for Efficient Chronic Disease Prediction
• On-Device Fine-Tuning via Backprop-Free Zeroth-Order Optimization
• Hindsight Distillation Reasoning with Knowledge Encouragement Preference for Knowledge-based Visual Question Answering
• Robustness of LLM-enabled vehicle trajectory prediction under data security threats
• Phantom Menace: Exploring and Enhancing the Robustness of VLA Models Against Physical Sensor Attacks
• ParoQuant: Pairwise Rotation Quantization for Efficient Reasoning LLM Inference
• Advanced Black-Box Tuning of Large Language Models with Limited API Calls
• Do Not Merge My Model! Safeguarding Open-Source LLMs Against Unauthorized Model Merging
• ChEmREF: Evaluating Language Model Readiness for Chemical Emergency Response
• Audio-VLA: Adding Contact Audio Perception to Vision-Language-Action Model for Robotic Manipulation
• EnchTable: Unified Safety Alignment Transfer in Fine-tuned Large Language Models
• Mastering Olympiad-Level Physics with Artificial Intelligence
• Modeling Uncertainty Trends for Timely Retrieval in Dynamic RAG
• SlideBot: A Multi-Agent Framework for Generating Informative, Reliable, Multi-Modal Presentations
• MCAD: Multimodal Context-Aware Audio Description Generation For Soccer
• MACEval: A Multi-Agent Continual Evaluation Network for Large Models
• Convergence and Stability Analysis of Self-Consuming Generative Models with Heterogeneous Human Curation
• TransactionGPT
• HalluClean: A Unified Framework to Combat Hallucinations in LLMs
• DPRM: A Dual Implicit Process Reward Model in Multi-Hop Question Answering
• AgentPRM: Process Reward Models for LLM Agents via Step-Wise Promise and Progress
• Remodeling Semantic Relationships in Vision-Language Fine-Tuning
• Relation as a Prior: A Novel Paradigm for LLM-based Document-level Relation Extraction
• A Small Leak Sinks All: Exploring the Transferable Vulnerability of Source Code Models
• HyCoRA: Hyper-Contrastive Role-Adaptive Learning for Role-Playing
• SERL: Self-Examining Reinforcement Learning on Open-Domain
• LoopLLM: Transferable Energy-Latency Attacks in LLMs via Repetitive Generation
• ViPRA: Video Prediction for Robot Actions
• Optimal Attention Temperature Enhances In-Context Learning under Distribution Shift
• Probabilities Are All You Need: A Probability-Only Approach to Uncertainty Estimation in Large Language Models
• Revisiting NLI: Towards Cost-Effective and Human-Aligned Metrics for Evaluating LLMs in Question Answering
• When, What, and How: Rethinking Retrieval-Enhanced Speculative Decoding
• Routing Manifold Alignment Improves Generalization of Mixture-of-Experts LLMs
• Voice-Interactive Surgical Agent for Multimodal Patient Data Control
• StreamKV: Streaming Video Question-Answering with Segment-based KV Cache Retrieval and Compression
• Synergy over Discrepancy: A Partition-Based Approach to Multi-Domain LLM Fine-Tuning
• GoCkpt: Gradient-Assisted Multi-Step overlapped Checkpointing for Efficient LLM Training
• HLPD: Aligning LLMs to Human Language Preference for Machine-Revised Text Detection
• Attention and Compression is all you need for Controllably Efficient Language Models
• LLMs vs. Traditional Sentiment Tools in Psychology: An Evaluation on Belgian-Dutch Narratives
• FedRW: Efficient Privacy-Preserving Data Reweighting for Enhancing Federated Learning of Language Models
• Ghost in the Transformer: Detecting Model Reuse with Invariant Spectral Signatures
• BuildingWorld: A Structured 3D Building Dataset for Urban Foundation Models
• EcoSpa: Efficient Transformer Training with Coupled Sparsity
• AdaDrive: Self-Adaptive Slow-Fast System for Language-Grounded Autonomous Driving
• Chain-of-Thought as a Lens: Evaluating Structured Reasoning Alignment between Human Preferences and Large Language Models
• TimeSense:Making Large Language Models Proficient in Time-Series Analysis
• 10 Open Challenges Steering the Future of Vision-Language-Action Models
• Interaction-Centric Knowledge Infusion and Transfer for Open-Vocabulary Scene Graph Generation
• VLDrive: Vision-Augmented Lightweight MLLMs for Efficient Language-grounded Autonomous Driving
• Can Fine-Tuning Erase Your Edits? On the Fragile Coexistence of Knowledge Editing and Adaptation
• DRAGON: Guard LLM Unlearning in Context via Negative Detection and Reasoning
• Steering Language Models with Weight Arithmetic
• EveryDayVLA: A Vision-Language-Action Model for Affordable Robotic Manipulation
• LiveStar: Live Streaming Assistant for Real-World Online Video Understanding
• 8bit-GPT: Exploring Human-AI Interaction on Obsolete Macintosh Operating Systems
• Linear Gradient Prediction with Control Variates
• Effectiveness of Chain-of-Thought in Distilling Reasoning Capability from Large Language Models
• Iterative Layer-wise Distillation for Efficient Compression of Large Language Models
• Scientific judgment drifts over time in AI ideation
• Leak@k: Unlearning Does Not Make LLMs Forget Under Probabilistic Decoding
• Personalized Image Editing in Text-to-Image Diffusion Models via Collaborative Direct Preference Optimization
• Cambrian-S: Towards Spatial Supersensing in Video
• An MLCommons Scientific Benchmarks Ontology
• From Model to Breach: Towards Actionable LLM-Generated Vulnerabilities Reporting
• Ground-Truth Subgraphs for Better Training and Evaluation of Knowledge Graph Augmented LLMs
• Advancing Equitable AI: Evaluating Cultural Expressiveness in LLMs for Latin American Contexts
• Plan of Knowledge: Retrieval-Augmented Large Language Models for Temporal Knowledge Graph Question Answering
• Memory- and Latency-Constrained Inference of Large Language Models via Adaptive Split Computing
• Direct Semantic Communication Between Large Language Models via Vector Translation
• Evaluating Modern Large Language Models on Low-Resource and Morphologically Rich Languages:A Cross-Lingual Benchmark Across Cantonese, Japanese, and Turkish
• ASVRI-Legal: Fine-Tuning LLMs with Retrieval Augmented Generation for Enhanced Legal Regulation
• From Prompts to Power: Measuring the Energy Footprint of LLM Inference
• Let the Bees Find the Weak Spots: A Path Planning Perspective on Multi-Turn Jailbreak Attacks against LLMs
• Automated Prompt Generation for Code Intelligence: An Empirical study and Experience in WeChat
• From Insight to Exploit: Leveraging LLM Collaboration for Adaptive Adversarial Text Generation
• Silenced Biases: The Dark Side LLMs Learned to Refuse
• Optimal Singular Damage: Efficient LLM Inference in Low Storage Regimes
• Can Visual Input Be Compressed? A Visual Token Compression Benchmark for Large Multimodal Models
• ECO Decoding: Entropy-Based Control for Controllability and Fluency in Controllable Dialogue Generation
• FP8-Flow-MoE: A Casting-Free FP8 Recipe without Double Quantization Error
• TabDSR: Decompose, Sanitize, and Reason for Complex Numerical Reasoning in Tabular Data
• Rethinking LLM Human Simulation: When a Graph is What You Need
• LLM Probing with Contrastive Eigenproblems: Improving Understanding and Applicability of CCS
• Rescuing the Unpoisoned: Efficient Defense against Knowledge Corruption Attacks on RAG Systems
• Improving Romanian LLM Pretraining Data using Diversity and Quality Filtering
• AI Progress Should Be Measured by Capability-Per-Resource, Not Scale Alone: A Framework for Gradient-Guided Resource Allocation in LLMs
• Efficient Test-Time Retrieval Augmented Generation
• ORANGE: An Online Reflection ANd GEneration framework with Domain Knowledge for Text-to-SQL
• G2rammar: Bilingual Grammar Modeling for Enhanced Text-attributed Graph Learning
• TriCon-Fair: Triplet Contrastive Learning for Mitigating Social Bias in Pre-trained Language Models
• Optimizing Native Sparse Attention with Latent Attention and Local Global Alternating Strategies
• Do You Know About My Nation? Investigating Multilingual Language Models' Cultural Literacy Through Factual Knowledge
• SpecDiff-2: Scaling Diffusion Drafter Alignment For Faster Speculative Decoding
• Efficiency vs. Alignment: Investigating Safety and Fairness Risks in Parameter-Efficient Fine-Tuning of LLMs
• Reimagining Safety Alignment with An Image
• SQLSpace: A Representation Space for Text-to-SQL to Discover and Mitigate Robustness Gaps
• Scalable Processing-Near-Memory for 1M-Token LLM Inference: CXL-Enabled KV-Cache Management Beyond GPU Limits
• SteerVLM: Robust Model Control through Lightweight Activation Steering for Vision Language Models
• LoRAQuant: Mixed-Precision Quantization of LoRA to Ultra-Low Bits
• Emu3.5: Native Multimodal Models are World Learners
• Envisioning Future Interactive Web Development: Editing Webpage with Natural Language
• Chain-of-Thought Hijacking
• MossNet: Mixture of State-Space Experts is a Multi-Head Attention
• Beyond Benchmarks: The Economics of AI Inference
• Loquetier: A Virtualized Multi-LoRA Framework for Unified LLM Fine-tuning and Serving
• AttnCache: Accelerating Self-Attention Inference for LLM Prefill via Attention Cache
• Instance-Level Composed Image Retrieval
• Gaperon: A Peppered English-French Generative Language Model Suite
• Robotic Assistant: Completing Collaborative Tasks with Dexterous Vision-Language-Action Models
• MODUS: Decoder-Only Any-to-Any Modeling of Diverse Modalities
• Mitigating privacy risks in Retrieval-Augmented Generation via locally private entity perturbation
• Towards accurate differential diagnosis with large language models
• Investigating the Validity Evidence of Automated Scoring Methods for Divergent Thinking Assessments
• Considering the ethics of large machine learning models in the chemical sciences
• Revisiting Lossy Verification in Speculative Decoding: Mechanisms, Trade-offs, and Failure Modes
• Relation Geometry in Semantic Space of Language Models
• Between Gradient and Natural Gradient: A Continuum of LoRA Initializations
• RedKnot: Efficient Long-Context LLM Serving with Head-Aware KV Reuse and SegPagedAttention
• Collaborating Foundation Models for Domain Generalized Semantic Segmentation
• Semantic Similarity in Radiology Reports via LLMs and NER
• Shades of zero: Distinguishing impossibility from inconceivability
• Why ‘open’ AI systems are actually closed, and why this matters
• Covenant-72B: Pre-Training a 72B LLM with Trustless Peers Over-the-Internet
• How do large-language models respond to moral dilemmas? Insights from the defining issues test
• Forecasting Trajectory-Level Safety Risks in Black-Box Multi-Turn Interactions
• Don't Just Chase "Highlighted Tokens" in MLLMs: Revisiting Visual Holistic Context Retention
• Can neurons speak? Semantic narration of vision at single-cell resolution
• Flows: Building Blocks of Reasoning and Collaborating AI
• Breaking the Ice: Analyzing Cold Start Latency in vLLM
• HRM-Text: Efficient Pretraining Beyond Scaling
• Back into Plato's Cave: Examining Cross-modal Representational Convergence at Scale
• CatPath‐GPT: A Mixture of Experts System for Computational Catalyst Design
• France or Spain or Germany or France: A Neural Account of Non-Redundant Redundant Disjunctions
• NeurIPT: Foundation Model for Neural Interfaces
• RL makes MLLMs see better than SFT
• EDVD-LLaMA: Explainable Deepfake Video Detection via Multimodal Large Language Model Reasoning
• State of the Art of LLM-Enabled Interaction with Visualization
• IBNorm: Information-Bottleneck Inspired Normalization for Representation Learning
• MoEntwine: Unleashing the Potential of Wafer-scale Chips for Large-scale Expert Parallel Inference
• Optimizing Knowledge Utilization for Multi-Intent Comment Generation with Large Language Models
• DTKG: Dual-Track Knowledge Graph-Verified Reasoning Framework for Multi-Hop QA
• NanoVLA: Routing Decoupled Vision-Language Understanding for Nano-sized Generalist Robotic Policies
• Opt4GPTQ: Co-Optimizing Memory and Computation for 4-bit GPTQ Quantized LLM Inference on Heterogeneous Platforms
• SemCoT: Accelerating Chain-of-Thought Reasoning through Semantically-Aligned Implicit Tokens
• Greedy Sampling Is Provably Efficient for RLHF
• Improving LLM Reasoning via Dependency-Aware Query Decomposition and Logic-Parallel Content Expansion
• LoRA-DA: Data-Aware Initialization for Low-Rank Adaptation via Asymptotic Analysis
• From Cross-Task Examples to In-Task Prompts: A Graph-Based Pseudo-Labeling Framework for In-context Learning
• LuxIT: A Luxembourgish Instruction Tuning Dataset from Monolingual Seed Data
• MiniOneRec: An Open-Source Framework for Scaling Generative Recommendation
• Critique-RL: Training Language Models for Critiquing through Two-Stage Reinforcement Learning
• Verifying Large Language Models' Reasoning Paths via Correlation Matrix Rank
• Evaluating LLMs on Generating Age-Appropriate Child-Like Conversations
• Kernelized Sparse Fine-Tuning with Bi-level Parameter Competition for Vision Models
• MeCeFO: Enhancing LLM Training Robustness via Fault-Tolerant Optimization
• PICOs-RAG: PICO-supported Query Rewriting for Retrieval-Augmented Generation in Evidence-Based Medicine
• Uncovering the Potential Risks in Unlearning: Danger of English-only Unlearning in Multilingual LLMs
• Agentic AI Security: Threats, Defenses, Evaluation, and Open Challenges
• Large Language Models, Agency, and Why Speech Acts are Beyond Them (For Now) – A Kantian-Cum-Pragmatist Case
• MoPHES:Leveraging on-device LLMs as Agent for Mobile Psychological Health Evaluation and Support
• MoS-VLA: A Vision-Language-Action Model with One-Shot Skill Adaptation
• ScaLoRA: Optimally Scaled Low-Rank Adaptation for Efficient High-Rank Fine-Tuning
• RoboOmni: Proactive Robot Manipulation in Omni-modal Context
• A Survey on Efficient Vision-Language-Action Models
• FRONTIER-RevRec: A Large-scale Dataset for Reviewer Recommendation
• Larger and more instructable language models become less reliable
• Dexbotic: Open-Source Vision-Language-Action Toolbox
• Emotion-Coherent Reasoning for Multimodal LLMs via Emotional Rationale Verifier
• How AI Forecasts AI Jobs: Benchmarking LLM Predictions of Labor Market Changes
• Nested AutoRegressive Models
• Simple Denoising Diffusion Language Models
• Can Language Models Compose Skills In-Context?
• Probing Knowledge Holes in Unlearned LLMs
• SeeDNorm: Self-Rescaled Dynamic Normalization
• CANDI: Hybrid Discrete-Continuous Diffusion Models
• Frustratingly Easy Task-aware Pruning for Large Language Models
• The Structural Scalpel: Automated Contiguous Layer Pruning for Large Language Models
• Label Smoothing Improves Gradient Ascent in LLM Unlearning
• GigaEmbeddings: Efficient Russian Language Embedding Model
• DETECT: Determining Ease and Textual Clarity of German Text Simplifications
• Mitigating Coordinate Prediction Bias from Positional Encoding Failures
• Sprint: Sparse-Dense Residual Fusion for Efficient Diffusion Transformers
• Brain-tuning Improves Generalizability and Efficiency of Brain Alignment in Speech Models
• FineRS: Fine-grained Reasoning and Segmentation of Small Objects with Reinforcement Learning
• MoniTor: Exploiting Large Language Models with Instruction for Online Video Anomaly Detection
• Large Language Models as Model Organisms for Human Associative Learning
• Flight Delay Prediction via Cross-Modality Adaptation of Large Language Models and Aircraft Trajectory Representation
• Evaluating Prompting Strategies and Large Language Models in Systematic Literature Review Screening: Relevance and Task-Stage Classification
• Bi-Level Optimization for Generative Recommendation: Bridging Tokenization and Generation
• Model Merging with Functional Dual Anchors
• The "Right" Discourse on Migration: Analysing Migration-Related Tweets in Right and Far-Right Political Movements
• Adjacent Words, Divergent Intents: Jailbreaking Large Language Models via Task Concurrency
• Large Language Models Meet Text-Attributed Graphs: A Survey of Integration Frameworks and Applications
• Robust Distortion-Free Watermark for Autoregressive Audio Generation Models
• Self-Rewarding PPO: Aligning Large Language Models with Demonstrations Only
• The Gray Zone of Faithfulness: Taming Ambiguity in Unfaithfulness Detection
• Stateful KV Cache Management for LLMs: Balancing Space, Time, Accuracy, and Positional Fidelity
• Out-of-distribution Tests Reveal Compositionality in Chess Transformers
• HA-RAG: Hotness-Aware RAG Acceleration via Mixed Precision and Data Placement
• HyperET: Efficient Training in Hyperbolic Space for Multi-modal Large Language Models
• Layer as Puzzle Pieces: Compressing Large Language Models through Layer Concatenation
• SAID: Safety-Aware Intent Defense via Prefix Probing for Large Language Models
• StableSketcher: Enhancing Diffusion Model for Pixel-based Sketch Generation via Visual Question Answering Feedback
• LM-mixup: Text Data Augmentation via Language Model based Mixup
• RECALL: REpresentation-aligned Catastrophic-forgetting ALLeviation via Hierarchical Model Merging
• The Impact of Negated Text on Hallucination with Large Language Models
• Do Prompts Reshape Representations? An Empirical Study of Prompting Effects on Embeddings
• LaViRA: Language-Vision-Robot Actions Translation for Zero-Shot Vision Language Navigation in Continuous Environments
• Energy-Efficient and Dequantization-Free Q-LLMs: A Spiking Neural Network Approach to Salient Value Mitigation
• Machine Text Detectors are Membership Inference Attacks
• KnowMol: Advancing Molecular Large Language Models with Multi-Level Chemical Knowledge
• ELUTQ: Efficient LUT-Aware Quantization for Deploying Large Language Models on Edge Devices
• LLM Unlearning with LLM Beliefs
• AgenticMath: Enhancing LLM Reasoning via Agentic-based Math Data Generation
• Conditions for Catastrophic Forgetting in Multilingual Translation
• Investigating LLM Capabilities on Long Context Comprehension for Medical Question Answering
• Pay Attention to the Triggers: Constructing Backdoors That Survive Distillation
• Simple and Efficient Heterogeneous Temporal Graph Neural Network
• From Retrieval to Generation: Unifying External and Parametric Knowledge for Medical Question Answering
• Any-Depth Alignment: Unlocking Innate Safety Alignment of LLMs to Any-Depth
• From Local to Global: Revisiting Structured Pruning Paradigms for Large Language Models
• Assessing Monotone Dependence: Area Under the Curve Meets Rank Correlation
• Foundational Automatic Evaluators: Scaling Multi-Task Generative Evaluator Training for Reasoning-Centric Domains
• SparseVILA: Decoupling Visual Sparsity for Efficient VLM Inference
• Voice EHR: introducing multimodal audio data for health
• Protein Language Models: Is Scaling Necessary?
• ReXMoE: Reusing Experts with Minimal Overhead in Mixture-of-Experts
• Explainability of Large Language Models: Opportunities and Challenges toward Generating Trustworthy Explanations
• Select-Then-Decompose: From Empirical Analysis to Adaptive Selection Strategy for Task Decomposition in Large Language Models
• Forget to Know, Remember to Use: Context-Aware Unlearning for Large Language Models
• Wisdom is Knowing What not to Say: Hallucination-Free LLMs Unlearning via Attention Shifting
• VisiPruner: Decoding Discontinuous Cross-Modal Dynamics for Efficient Multimodal LLMs
• Efficient Vision-Language-Action Models for Embodied Manipulation: A Systematic Survey
• Mapping from Meaning: Addressing the Miscalibration of Prompt-Sensitive Language Models
• Forgetting to Forget: Attention Sink as A Gateway for Backdooring LLM Unlearning
• A Comprehensive Survey on Reinforcement Learning-based Agentic Search: Foundations, Roles, Optimizations, Evaluations, and Applications
• QuanBench: Benchmarking Quantum Code Generation with Large Language Models
• MuonBP: Faster Muon via Block-Periodic Orthogonalization
• EMRRG: Efficient Fine-Tuning Pre-trained X-ray Mamba Networks for Radiology Report Generation
• Advances in Pre-trained Language Models for Domain-Specific Text Classification: A Systematic Review
• Accelerating Mobile Language Model via Speculative Decoding and NPU-Coordinated Execution
• AUGUSTUS: An LLM-Driven Multimodal Agent System with Contextualized User Memory
• InfiMed-ORBIT: Aligning LLMs on Open-Ended Complex Tasks via Rubric-Based Incremental Training
• Teaching Code LLMs to Use Autocompletion Tools in Repository-Level Code Generation
• Enhancing reliability in AI inference services: An empirical study on real production incidents
• Synera: Synergistic LLM Serving across Device and Cloud at Scale
• How Well Can Preference Optimization Generalize Under Noisy Feedback?
• Train a Unified Multimodal Data Quality Classifier with Synthetic Data
• From Pixels to Words -- Towards Native Vision-Language Primitives at Scale
• Attention Is All You Need for KV Cache in Diffusion LLMs
• Predicting Task Performance with Context-aware Scaling Laws
• Purifying Task Vectors in Knowledge-Aware Subspace for Model Merging
• Can MLLMs Absorb Math Reasoning Abilities from LLMs as Free Lunch?
• Stop-RAG: Value-Based Retrieval Control for Iterative RAG
• Large Reasoning Embedding Models: Towards Next-Generation Dense Retrieval Paradigm
• MERLIN: A Testbed for Multilingual Multimodal Entity Recognition and Linking
• Kelle: Co-design KV Caching and eDRAM for Efficient LLM Serving in Edge Computing
• FedHFT: Efficient Federated Finetuning with Heterogeneous Edge Clients
• Think Globally, Group Locally: Evaluating LLMs Using Multi-Lingual Word Grouping Games
• MemoTime: Memory-Augmented Temporal Knowledge Graph Enhanced Large Language Model Reasoning
• Document Intelligence in the Era of Large Language Models: A Survey
• End-to-End Multi-Modal Diffusion Mamba
• Synthesizing Agentic Data for Web Agents with Progressive Difficulty Enhancement Mechanisms
• Information-Theoretic Reward Modeling for Stable RLHF: Detecting and Mitigating Reward Hacking
• Tahakom LLM Guidelines and Recipes: From Pre-training Data to an Arabic LLM
• Taming the Fragility of KV Cache Eviction in LLM Inference
• UniMoE-Audio: Unified Speech and Music Generation with Dynamic-Capacity MoE
• TRUSTVIS: A Multi-Dimensional Trustworthiness Evaluation Framework for Large Language Models
• MedREK: Retrieval-Based Editing for Medical LLMs with Key-Aware Prompts
• Beyond Static LLM Policies: Imitation-Enhanced Reinforcement Learning for Recommendation
• ConsintBench: Evaluating Language Models on Real-World Consumer Intent Understanding
• Hierarchical Frequency Tagging Probe (HFTP): A Unified Approach to Investigate Syntactic Structure Representations in Large Language Models and the Human Brain
• D-com: Accelerating Iterative Processing to Enable Low-rank Decomposition of Activations
• Repairing Reward Functions with Human Feedback to Mitigate Reward Hacking
• Litespark Technical Report: High-Throughput, Energy-Efficient LLM Training Framework
• RAID: Refusal-Aware and Integrated Decoding for Jailbreaking LLMs
• Cautious Weight Decay
• Pretraining in Actor-Critic Reinforcement Learning for Robot Locomotion
• Traveling Salesman-Based Token Ordering Improves Stability in Homomorphically Encrypted Language Models
• HiLoRA: Adaptive Hierarchical LoRA Routing for Training-Free Domain Generalization
• MoRA: On-the-fly Molecule-aware Low-Rank Adaptation Framework for LLM-based Multi-Modal Molecular Assistant
• Unveiling the Vulnerability of Graph-LLMs: An Interpretable Multi-Dimensional Adversarial Attack on TAGs
• Evolution of meta's llama models and parameter-efficient fine-tuning of large language models: a survey
• A Survey on Parallel Reasoning
• Precise Attribute Intensity Control in Large Language Models via Targeted Representation Editing
• Empowering LLM Agents with Geospatial Awareness: Toward Grounded Reasoning for Wildfire Response
• Hierarchical Alignment: Surgical Fine-Tuning via Functional Layer Specialization in Large Language Models
• CPR: Mitigating Large Language Model Hallucinations with Curative Prompt Refinement
• SVAG-Bench: A Large-Scale Benchmark for Multi-Instance Spatio-temporal Video Action Grounding
• OPLoRA: Orthogonal Projection LoRA Prevents Catastrophic Forgetting during Parameter-Efficient Fine-Tuning
• Spatial Forcing: Implicit Spatial Representation Alignment for Vision-language-action Model
• Multi-stage Prompt Refinement for Mitigating Hallucinations in Large Language Models
• FinVet: A Collaborative Framework of RAG and External Fact-Checking Agents for Financial Misinformation Detection
• Ontolearn-A Framework for Large-scale OWL Class Expression Learning in Python
• Hallucination Detection via Internal States and Structured Reasoning Consistency in Large Language Models
• Diffusion-Link: Diffusion Probabilistic Model for Bridging the Audio-Text Modality Gap
• Neural Weight Compression for Language Models
• CoDefend: Cross-Modal Collaborative Defense via Diffusion Purification and Prompt Optimization
• Bolster Hallucination Detection via Prompt-Guided Data Augmentation
• AccurateRAG: A Framework for Building Accurate Retrieval-Augmented Question-Answering Applications
• xLSTM Scaling Laws: Competitive Performance with Linear Time-Complexity
• AMiD: Knowledge Distillation for LLMs with α-mixture Assistant Distribution
• Preserving LLM Capabilities through Calibration Data Curation: From Analysis to Optimization
• SASER: Stego attacks on open-source LLMs
• Long Exposure: Accelerating Parameter-Efficient Fine-Tuning for LLMs under Shadowy Sparsity
• Unlocking LLM Safeguards for Low-Resource Languages via Reasoning and Alignment with Minimal Training Data
• DynaSpec: Context-aware Dynamic Speculative Sampling for Large-Vocabulary Language Models
• MetaBreak: Jailbreaking Online LLM Services via Special Token Manipulation
• Debiasing LLMs by Masking Unfairness-Driving Attention Heads
• PermLLM: Learnable Channel Permutation for N:M Sparse Large Language Models
• Diversity Augmentation of Dynamic User Preference Data for Boosting Personalized Text Summarizers
• Improving Speech Emotion Recognition with Mutual Information Regularized Generative Model
• Automated Glaucoma Report Generation via Dual-Attention Semantic Parallel-LSTM and Multimodal Clinical Data Integration
• StelLA: Subspace Learning in Low-rank Adaptation using Stiefel Manifold
• Detecting LLM-Generated Spam Reviews by Integrating Language Model Embeddings and Graph Neural Network
• Do Thought Streams Matter? Evaluating Reasoning in Gemini Vision-Language Models for Video Scene Understanding
• Understanding and Exploiting Weight Update Sparsity for Communication-Efficient Distributed RL
• Benchmarking Language Modeling for Lossless Compression of Full-Fidelity Audio
• CoIDO: Efficient Data Selection for Visual Instruction Tuning via Coupled Importance-Diversity Optimization
• Steering Embedding Models with Geometric Rotation: Mapping Semantic Relationships Across Languages and Models
• SIMBA UQ: Similarity-Based Aggregation for Uncertainty Quantification in Large Language Models
• FLRC: Fine-grained Low-Rank Compressor for Efficient LLM Inference
• Entropy Meets Importance: A Unified Head Importance-Entropy Score for Stable and Efficient Transformer Pruning
• RoDyn: Taming Interactive Robot-Dynamic 2.5D World Model for Robotic Manipulation
• All Code, No Thought: Current Language Models Struggle to Reason in Ciphered Language
• Towards Better & Faster Autoregressive Image Generation: From the Perspective of Entropy
• LLM Unlearning on Noisy Forget Sets: A Study of Incomplete, Rewritten, and Watermarked Data
• Bridging the Semantic Gap: Contrastive Rewards for Multilingual Text-to-SQL with GRPO
• Score-Based Density Estimation from Pairwise Comparisons
• Goal-oriented Backdoor Attack against Vision-Language-Action Models via Physical Objects
• Active Model Selection for Large Language Models
• Q-Router: Agentic Video Quality Assessment with Expert Model Routing and Artifact Localization
• AI-generated text boundary detection with RoFT
• Revisiting Hallucination Detection with Effective Rank-based Uncertainty
• Mitigating Judgment Preference Bias in Large Language Models through Group-Based Polling
• From Defender to Devil? Unintended Risk Interactions Induced by LLM Defenses
• LightReasoner: Can Small Language Models Teach Large Language Models Reasoning?
• MetaDefense: Defending Finetuning-based Jailbreak Attack Before and During Generation
• OBCache: Optimal Brain KV Cache Pruning for Efficient Long-Context LLM Inference
• RCPU: Rotation-Constrained Error Compensation for Structured Pruning of Large Language Models
• AutoRed: A Free-form Adversarial Prompt Generation Framework for Automated Red Teaming
• VisualDAN: Exposing Vulnerabilities in VLMs with Visual-Driven DAN Commands
• Don't Run with Scissors: Pruning Breaks VLA Models but They Can Be Recovered
• Learning What to Remember: Adaptive Probabilistic Memory Retention for Memory-Efficient Language Models
• TRIM: Token-wise Attention-Derived Saliency for Data-Efficient Instruction Tuning
• SCOPE-RL: Stable and Quantitative Control of Policy Entropy in RL Post-Training
• RetouchLLM: Training-free Code-based Image Retouching with Vision Language Models
• MLLM4TS: Leveraging Vision and Multimodal Language Models for General Time-Series Analysis
• Black-Box Detection of LLM-Generated Text Using Generalized Jensen-Shannon Divergence
• Red-Bandit: Test-Time Adaptation for LLM Red-Teaming via Bandit-Guided LoRA Experts
• Secure-Instruct: An Automated Pipeline for Synthesizing Instruction-Tuning Datasets Using LLMs for Secure Code Generation
• Vision-Language-Action Models for Robotics: A Review Towards Real-World Applications
• SHANKS: Simultaneous Hearing and Thinking for Spoken Language Models
• Unlocking Latent Discourse Translation in LLMs Through Quality-Aware Decoding
• SoftMatcha 2: A Fast and Soft Pattern Matcher for Trillion-Scale Corpora
• Growing Visual Generative Capacity for Pre-Trained MLLMs
• Mid-Training of Large Language Models: A Survey
• AWM: Accurate Weight-Matrix Fingerprint for Large Language Models
• PIKA: Expert-Level Synthetic Datasets for Post-Training Alignment from Scratch
• Reusing Overtrained Language Models Saturates Scaling
• Optimal Stopping vs Best-of-N for Inference Time Optimization
• Learning to Route LLMs from Bandit Feedback: One Policy, Many Trade-offs
• POME: Post Optimization Model Edit via Muon-style Projection
• Reward Model Perspectives: Whose Opinions Do Reward Models Reward?
• Leveraging Large Language Models for Cybersecurity Risk Assessment -- A Case from Forestry Cyber-Physical Systems
• TokenChain: A Discrete Speech Chain via Semantic Token Modeling
• Prompt reinforcing for long-term planning of large language models
• Mixture of Neuron Experts
• Primal-Dual Direct Preference Optimization for Constrained LLM Alignment
• Activation-Informed Pareto-Guided Low-Rank Compression for Efficient LLM/VLM
• ARMOR: High-Performance Semi-Structured Pruning via Adaptive Matrix Factorization
• From Principles to Practice: A Systematic Study of LLM Serving on Multi-core NPUs
• SD-MVSum: Script-Driven Multimodal Video Summarization Method and Datasets
• Latent Speech-Text Transformer
• The New Quant: A Survey of Large Language Models in Financial Prediction and Trading
• CAM: A Constructivist View of Agentic Memory for LLM-Based Reading Comprehension
• InvThink: Premortem Reasoning for Safer Language Models
• Test-Time Scaling in Diffusion LLMs via Hidden Semi-Autoregressive Experts
• Imperceptible Jailbreaking against Large Language Models
• HyperVLA: Efficient Inference in Vision-Language-Action Models via Hypernetworks
• Reproducibility Study of "XRec: Large Language Models for Explainable Recommendation"
• Demystifying Synthetic Data in LLM Pre-training: A Systematic Study of Scaling Laws, Benefits, and Pitfalls
• Beyond the Seen: Bounded Distribution Estimation for Open-Vocabulary Learning
• QuantAgents: Towards Multi-agent Financial System via Simulated Trading
• Compressed Convolutional Attention: Efficient Attention in a Compressed Latent Space
• KVLinC : KV Cache Quantization with Hadamard Rotation and Linear Correction
• Representation Potentials of Foundation Models for Multimodal Alignment: A Survey
• Evaluation of Clinical Trials Reporting Quality using Large Language Models
• LLM Microscope: What Model Internals Reveal About Answer Correctness and Context Utilization
• SECA: Semantically Equivalent and Coherent Attacks for Eliciting LLM Hallucinations
• The Unseen Frontier: Pushing the Limits of LLM Sparsity with Surrogate-Free ADMM
• Mechanistic Interpretability of Socio-Political Frames in Language Models
• Towards Sampling Data Structures for Tensor Products in Turnstile Streams
• Best-of-Majority: Minimax-Optimal Strategy for Pass@k Inference Scaling
• Knowledge Graph-Guided Multi-Agent Distillation for Reliable Industrial Question Answering with Datasets
• HALO: Memory-Centric Heterogeneous Accelerator with 2.5D Integration for Low-Batch LLM Inference
• Fine-Tuning Jailbreaks under Highly Constrained Black-Box Settings: A Three-Pronged Approach
• Benchmarking Foundation Models with Retrieval-Augmented Generation in Olympic-Level Physics Problem Solving
• Downgrade to Upgrade: Optimizer Simplification Enhances Robustness in LLM Unlearning
• HAMLET: Switch your Vision-Language-Action Model into a History-Aware Policy
• MASH: Modeling Abstention via Selective Help-Seeking
• Characterizing Model Behavior Under Synthetic Data Training: An Empirical Study Across Scales and Mixing Ratios
• Automated Structured Radiology Report Generation with Rich Clinical Context
• ZQBA: Zero Query Black-box Adversarial Attack
• Facilitating Cognitive Accessibility with LLMs: A Multi-Task Approach to Easy-to-Read Text Generation
• Direct Token Optimization: A Self-contained Approach to Large Language Model Unlearning
• Efficient Multi-modal Large Language Models via Progressive Consistency Distillation
• VLA-RFT: Vision-Language-Action Reinforcement Fine-tuning with Verified Rewards in World Simulators
• HiSpec: Hierarchical Speculative Decoding for LLMs
• Debunk the Myth of SFT Generalization
• Communication-Efficient and Accurate Approach for Aggregation in Federated Low-Rank Adaptation
• SafeBehavior: Simulating Human-Like Multistage Reasoning to Mitigate Jailbreak Attacks in Large Language Models
• Latent Thinking Optimization: Your Latent Reasoning Language Model Secretly Encodes Reward Signals in Its Latent Thoughts
• Explaining novel senses using definition generation with open language models
• Human-MME: A Holistic Evaluation Benchmark for Human-Centric Multimodal Large Language Models
• OWL: Geometry-Aware Spatial Reasoning for Audio Large Language Models
• CAST: Continuous and Differentiable Semi-Structured Sparsity-Aware Training for Large Language Models
• Reliability Crisis of Reference-free Metrics for Grammatical Error Correction
• Understanding the Mixture-of-Experts with Nadaraya-Watson Kernel
• SAIL: SRAM-Accelerated LLM Inference System with Lookup-Table-based GEMV
• TruthRL: Incentivizing Truthful LLMs via Reinforcement Learning
• Rotation Control Unlearning: Quantifying and Controlling Continuous Unlearning for LLM with The Cognitive Rotation Space
• Effective Model Pruning
• FlowMoE: A Scalable Pipeline Scheduling Framework for Distributed Mixture-of-Experts Training
• VLM-FO1: Bridging the Gap Between High-Level Reasoning and Fine-Grained Perception in VLMs
• Judging by Appearances? Auditing and Intervening Vision-Language Models for Bail Prediction
• VISOR++: Universal Visual Inputs based Steering for Large Vision Language Models
• Rethinking Parameter Sharing for LLM Fine-Tuning with Multiple LoRAs
• Rethinking Entropy Regularization in Large Reasoning Models
• Scaling with Collapse: Efficient and Predictable Training of LLM Families
• CLPO: Curriculum Learning meets Policy Optimization for LLM Reasoning
• MobileLLM-R1: Exploring the Limits of Sub-Billion Language Model Reasoners with Open Training Recipes
• Case reports unlocked: Harnessing large language models to advance research on child maltreatment
• Vision Function Layer in Multimodal LLMs
• Fidel-TS: A High-Fidelity Benchmark for Multimodal Time Series Forecasting
• Neural Message-Passing on Attention Graphs for Hallucination Detection
• Understanding the Dilemma of Unlearning for Large Language Models
• SparseServe: Unlocking Parallelism for Dynamic Sparse Attention in Long-Context LLM Serving
• Knowledge Editing with Subspace-Aware Key-Value Mappings
• Sanitize Your Responses: Mitigating Privacy Leakage in Large Language Models
• HarmMetric Eval: Benchmarking Metrics and Judges for LLM Harmfulness Assessment
• Evolution Strategies at Scale: LLM Fine-Tuning Beyond Reinforcement Learning
• AlignX: Advancing Multilingual Large Language Models with Multilingual Representation Alignment
• Speculative Verification: Exploiting Information Gain to Refine Speculative Decoding
• Multimodal Large Language Models Meet Multimodal Emotion Recognition and Reasoning: A Survey
• Exploring Similarity between Neural and LLM Trajectories in Language Processing
• Learning to Ponder: Adaptive Reasoning in Latent Space
• Conda: Column-Normalized Adam for Training Large Language Models Faster
• FM-FoG: A Real-Time Foundation Model-based Wearable System for Freezing-of-Gait Mitigation
• World-Env: Leveraging World Model as a Virtual Environment for VLA Post-Training
• Mitigating Visual Hallucinations via Semantic Curriculum Preference Optimization in MLLMs
• Beyond Repetition: Text Simplification and Curriculum Learning for Data-Constrained Pretraining
• Which course? Discourse! Teaching Discourse and Generation in the Era of LLMs
• Edge-FIT: Federated Instruction Tuning of Quantized LLMs for Privacy-Preserving Smart Home Environments
• Sequential Diffusion Language Models
• Detecting and Rectifying Noisy Labels: A Similarity-based Approach
• Dynamic Orthogonal Continual Fine-tuning for Mitigating Catastrophic Forgettings
• Winning the Pruning Gamble: A Unified Approach to Joint Sample and Token Pruning for Efficient Supervised Fine-Tuning
• Tequila: Trapping-free Ternary Quantization for Large Language Models
• Anchored Supervised Fine-Tuning
• AdaPtis: Reducing Pipeline Bubbles with Adaptive Pipeline Parallelism on Heterogeneous Models
• Beyond English-Centric Training: How Reinforcement Learning Improves Cross-Lingual Reasoning in LLMs
• Don't Settle Too Early: Self-Reflective Remasking for Diffusion Language Models
• MotionVerse: A Unified Multimodal Framework for Motion Comprehension, Generation and Editing
• Toward a Holistic Approach to Continual Model Merging
• LLM Hallucination Detection: HSAD
• Internal Planning in Language Models: Characterizing Horizon and Branch Awareness
• Clean First, Align Later: Benchmarking Preference Data Cleaning for Reliable LLM Alignment
• RIV: Recursive Introspection Mask Diffusion Vision Language Model
• QuadEnhancer: Leveraging Quadratic Transformations to Enhance Deep Neural Networks
• Dual-Space Smoothness for Robust and Balanced LLM Unlearning
• p-less Sampling: A Robust Hyperparameter-Free Approach for LLM Decoding
• Decision Potential Surface: A Theoretical and Practical Approximation of LLM's Decision Boundary
• General Exploratory Bonus for Optimistic Exploration in RLHF
• PT²-LLM: Post-Training Ternarization for Large Language Models
• Boltz-1 Democratizing Biomolecular Interaction Modeling
• Adaptive Margin RLHF via Preference over Preferences
• Memory-Efficient Fine-Tuning via Low-Rank Activation Compression
• Fine-Grained Detection of Context-Grounded Hallucinations Using LLMs
• Language Models Can Learn from Verbal Feedback Without Scalar Rewards
• Variational Reasoning for Language Models
• Your RAG is Unfair: Exposing Fairness Vulnerabilities in Retrieval-Augmented Generation via Backdoor Attacks
• Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
• Beyond Textual Context: Structural Graph Encoding with Adaptive Space Alignment to alleviate the hallucination of LLMs
• Bridging Draft Policy Misalignment: Group Tree Optimization for Speculative Decoding
• Multilingual Vision-Language Models, A Survey
• Mind the Missing: Variable-Aware Representation Learning for Irregular EHR Time Series using Large Language Models
• Action-aware Dynamic Pruning for Efficient Vision-Language-Action Manipulation
• COSPADI: Compressing LLMs via Calibration-Guided Sparse Dictionary Learning
• A2R: An Asymmetric Two-Stage Reasoning Framework for Parallel Reasoning
• Effect of Model Merging in Domain-Specific Ad-hoc Retrieval
• Elastic MoE: Unlocking the Inference-Time Scalability of Mixture-of-Experts
• Enhancing Low-Rank Adaptation with Structured Nonlinear Transformations
• SoK: Potentials and Challenges of Large Language Models for Reverse Engineering
• Backdoor Attribution: Elucidating and Controlling Backdoor in Language Models
• Where Did It Go Wrong? Attributing Undesirable LLM Behaviors via Representation Gradient Tracing
• Learning Admissible Heuristics for A*: Theory and Practice
• We Think, Therefore We Align LLMs to Helpful, Harmless and Honest Before They Go Wrong
• Compute-Optimal Quantization-Aware Training
• Towards Transparent AI: A Survey on Explainable Language Models
• Preemptive Detection and Steering of LLM Misalignment via Latent Reachability
• VideoJudge: Bootstrapping Enables Scalable Supervision of MLLM-as-a-Judge for Video Understanding
• VC-Agent: An Interactive Agent for Customized Video Dataset Collection
• Data-Centric Elastic Pipeline Parallelism for Efficient Long-Context LLM Training
• PMark: Towards Robust and Distortion-free Semantic-level Watermarking with Channel Constraints
• FORCE: Transferable Visual Jailbreaking Attacks via Feature Over-Reliance CorrEction
• CLUE: Conflict-guided Localization for LLM Unlearning Framework
• A short survey on almost orthogonal vectors in a few specific large dimensions
• ArchGPT: Understanding the World's Architectures with Large Multimodal Models
• Towards Atoms of Large Language Models
• SFT Doesn't Always Hurt General Capabilities: Revisiting Domain-Specific Fine-Tuning in LLMs
• IoT-MCP: Bridging LLMs and IoT Systems Through Model Context Protocol
• Beyond Stars: Bridging the Gap Between Ratings and Review Sentiment with LLM
• It's Not You, It's Clipping: A Soft Trust-Region via Probability Smoothing for LLM RL
• DeFacto: Counterfactual Thinking with Images for Enforcing Evidence-Grounded and Faithful Reasoning
• Building Tailored Speech Recognizers for Japanese Speaking Assessment
• LayoutAgent: A Vision-Language Agent Guided Compositional Diffusion for Spatial Layout Planning
• Instruction Boundary: Quantifying Biases in LLM Reasoning under Various Coverage
• Beyond Sharp Minima: Robust LLM Unlearning via Feedback-Guided Multi-Point Optimization
• Q-Palette: Fractional-Bit Quantizers Toward Optimal Bit Allocation for Efficient LLM Deployment
• BurstEngine: an Efficient Distributed Framework for Training Transformers on Extremely Long Sequences of over 1M Tokens
• Exploration with Foundation Models: Capabilities, Limitations, and Hybrid Approaches
• SSTAG: Structure-Aware Self-Supervised Learning Method for Text-Attributed Graphs
• CoMelSinger: Discrete Token-Based Zero-Shot Singing Synthesis With Structured Melody Control and Guidance
• FreezeVLA: Action-Freezing Attacks against Vision-Language-Action Models
• RSAVQ: Riemannian Sensitivity-Aware Vector Quantization for Large Language Models
• RoboSSM: Scalable In-context Imitation Learning via State-Space Models
• The future of machine learning for small-molecule drug discovery will be driven by data
• LOCA: Logical Chain Augmentation for Scientific Corpus Cleaning
• Detoxifying Large Language Models via Autoregressive Reward Guided Representation Editing
• bi-GRPO: Bidirectional Optimization for Jailbreak Backdoor Injection on LLMs
• Are We Scaling the Right Thing? A System Perspective on Test-Time Scaling
• Silent Tokens, Loud Effects: Padding in LLMs
• Parameter-Efficient Multi-Task Learning via Progressive Task-Specific Adaptation
• Harnessing the Potential of Optimizing Data Mixtures via Bayesian Domain Reweighting
• LightRot: A Light-Weighted Rotation Scheme and Architecture for Accurate Low-Bit Large Language Model Inference
• GyRot: Leveraging Hidden Synergy between Rotation and Fine-grained Group Quantization for Low-bit LLM Inference
• Comparing Llama-2 and GPT-3 LLMs for HPC kernels generation
• Conflicting feelings: sociological and computational sentiment in workplace sentiment surveillance
• Training Compute-Optimal Protein Language Models
• A semantic embedding space based on large language models for modelling human beliefs
• A large-scale evaluation of commonsense knowledge in humans and large language models
• Linguistic properties and model scale in brain encoding: from small to compressed language models
• Artificial intelligence-based biomarkers for treatment decisions in oncology
• Had Enough of Experts? Quantitative Knowledge Retrieval From Large Language Models
• Security and Privacy Challenges of Large Language Models: A Survey
• Understanding Subword Compositionality of Large Language Models
• How Do LLM-Generated Texts Impact Term-Based Retrieval Models?
• LLMulator: Generalizable Cost Modeling for Dataflow Accelerators with Input-Adaptive Control Flow
• A Knowledge Graph and a Tripartite Evaluation Framework Make Retrieval-Augmented Generation Scalable and Transparent
• CR-Net: Scaling Parameter-Efficient Training with Cross-Layer Low-Rank Structure
• LLM-based Agents Suffer from Hallucinations: A Survey of Taxonomy, Methods, and Directions
• Consistency-Aware Parameter-Preserving Knowledge Editing Framework for Multi-Hop Question Answering
• COLT: Enhancing Video Large Language Models with Continual Tool Usage
• Enhancing LLM-Based Social Bot via an Adversarial Learning Framework
• OraPO: Oracle-educated Reinforcement Learning for Data-efficient and Factual Radiology Report Generation
• Experience Scaling: Post-Deployment Evolution For Large Language Models
• CoCoA: Confidence and Context-Aware Adaptive Decoding for Resolving Knowledge Conflicts in Large Language Models
• Code Driven Planning with Domain-Adaptive Critic
• Advances in Large Language Models for Medicine
• Confidential LLM Inference: Performance and Cost Across CPU and GPU TEEs
• Symphony-MoE: Harmonizing Disparate Pre-trained Models into a Coherent Mixture-of-Experts
• RL-Finetuned LLMs for Privacy-Preserving Synthetic Rewriting
• Bounded PCTL Model Checking of Large Language Model Outputs
• Model selection meets clinical semantics: Optimizing ICD-10-CM prediction via LLM-as-Judge evaluation, redundancy-aware sampling, and section-aware fine-tuning
• Speculate Deep and Accurate: Lossless and Training-Free Acceleration for Offloaded LLMs via Substitute Speculative Decoding
• On-the-Fly Adaptation to Quantization: Configuration-Aware LoRA for Efficient Fine-Tuning of Quantized LLMs
• Enhancing Cross-Lingual Transfer through Reversible Transliteration: A Huffman-Based Approach for Low-Resource Languages
• Correlation or Causation: Analyzing the Causal Structures of LLM and LRM Reasoning Process
• Cronus: Efficient LLM inference on Heterogeneous GPU Clusters via Partially Disaggregated Prefill
• LLaVul: A Multimodal LLM for Interpretable Vulnerability Reasoning about Source Code
• Weights-Rotated Preference Optimization for Large Language Models
• MLLM-Driven Semantic Identifier Generation for Generative Cross-Modal Retrieval
• Training-free Truthfulness Detection via Sparse MLP Value Vectors
• DIWALI: Diversity and Inclusivity aWare cuLture specific Items for India: Dataset and Assessment of LLMs for Cultural Text Adaptation in Indian Context
• A State-Update Prompting Strategy for Efficient and Robust Multi-turn Dialogue
• PG-CE: A Progressive Generation Dataset with Constraint Enhancement for Controllable Text Generation
• Qwen3-Omni Technical Report
• Probabilistic Token Alignment for Large Language Model Fusion
• SnipSnap: A Joint Compression Format and Dataflow Co-Optimization Framework for Efficient Sparse LLM Accelerator Design
• CUTE: A Multilingual Dataset for Enhancing Cross-Lingual Knowledge Transfer in Low-Resource Languages
• PTQTP: Post-Training Quantization to Trit-Planes for Large Language Models
• Uncovering Implicit Bias in Large Language Models with Concept Learning Dataset
• ExpertWeave: Efficiently Serving Expert-Specialized Fine-Tuned Adapters at Scale
• Steering When Necessary: Flexible Steering Large Language Models with Backtracking
• Catching the Details: Self-Distilled RoI Predictors for Fine-Grained MLLM Perception
• Adaptive Graph Convolution and Semantic-Guided Attention for Multimodal Risk Detection in Social Networks
• DecipherGuard: Understanding and Deciphering Jailbreak Prompts for a Safer Deployment of Intelligent Software Systems
• Sycophancy Mitigation Through Reinforcement Learning with Uncertainty-Aware Adaptive Reasoning Trajectories
• Decoding Uncertainty: The Impact of Decoding Strategies for Uncertainty Estimation in Large Language Models
• MCP: A Control-Theoretic Orchestration Framework for Synergistic Efficiency and Interpretability in Multimodal Large Language Models
• Analyzing the Effects of Supervised Fine-Tuning on Model Knowledge from Token and Parameter Levels
• LLMsPark: A Benchmark for Evaluating Large Language Models in Strategic Gaming Contexts
• Can an Individual Manipulate the Collective Decisions of Multi-Agents?
• Rethinking the Role of Text Complexity in Language Model Pretraining
• HERO: Hierarchical Extrapolation and Refresh for Efficient World Models
• On Optimal Steering to Achieve Exact Fairness
• Pipeline Parallelism is All You Need for Optimized Early-Exit Based Self-Speculative Decoding
• LiteLong: Resource-Efficient Long-Context Data Synthesis for LLMs
• Brevity is the Soul of Wit: Condensing Code Changes to Improve Commit Message Generation
• On the Convergence of Muon and Beyond
• Generalizability of Large Language Model-Based Agents: A Comprehensive Survey
• DivLogicEval: A Framework for Benchmarking Logical Reasoning Evaluation in Large Language Models
• LNE-Blocking: An Efficient Framework for Contamination Mitigation Evaluation on Large Language Models
• Understand Before You Generate: Self-Guided Training for Autoregressive Image Generation
• Semantic Representation Attack against Aligned Large Language Models
• From Hype to Insight: Rethinking Large Language Model Integration in Visual Speech Recognition
• Towards Building Speech Large Language Models for Multitask Understanding in Low-Resource Languages
• LEAP: LLM Inference on Scalable PIM-NoC Architecture with Balanced Dataflow and Fine-Grained Parallelism
• Evaluating Large Language Models for Cross-Lingual Retrieval
• Decoupled Proxy Alignment: Mitigating Language Prior Conflict for Multimodal Alignment in MLLM
• Controlling Language Difficulty in Dialogues with Linguistic Features
• LLM Jailbreak Detection for (Almost) Free!
• Adaptive LoRA Experts Allocation and Selection for Federated Fine-Tuning
• OpenViGA: Video Generation for Automotive Driving Scenes by Streamlining and Fine-Tuning Open Source Models with Public Data
• TableDART: Dynamic Adaptive Multi-Modal Routing for Table Understanding
• Do LLMs Align Human Values Regarding Social Biases? Judging and Explaining Social Biases with LLMs
• UI-Level Evaluation of ALLaM 34B: Measuring an Arabic-Centric LLM via HUMAIN Chat
• ST-LINK: Spatially-Aware Large Language Models for Spatio-Temporal Forecasting
• Q-ROAR: Outlier-Aware Rescaling for RoPE Position Interpolation in Quantized Long-Context LLMs
• Keywords are not always the key: A metadata field analysis for natural language search on open data portals
• DSCC-HS: A Dynamic Self-Reinforcing Framework for Hallucination Suppression in Large Language Models
• Hala Technical Report: Building Arabic-Centric Instruction & Translation Models at Scale
• Do Large Language Models Understand Word Senses?
• Programmable Cognitive Bias in Social Agents
• Divergences between Language Models and Human Brains
• The Few-shot Dilemma: Over-prompting Large Language Models
• HPIM: Heterogeneous Processing-In-Memory-based Accelerator for Large Language Models Inference
• EvoEmpirBench: Dynamic Spatial Reasoning with Agent-ExpVer
• Beyond Artificial Misalignment: Detecting and Grounding Semantic-Coordinated Multimodal Manipulations
• A Systematic Evaluation of Parameter-Efficient Fine-Tuning Methods for the Security of Code LLMs
• Empowering LLMs with Parameterized Skills for Adversarial Long-Horizon Planning
• Module-Aware Parameter-Efficient Machine Unlearning on Transformers
• SSFO: Self-Supervised Faithfulness Optimization for Retrieval-Augmented Generation
• Memory-Efficient Federated Fine-Tuning of Large Language Models via Layer Pruning
• InfoGain-RAG: Boosting Retrieval-Augmented Generation via Document Information Gain-based Reranking and Filtering
• Towards Better Correctness and Efficiency in Code Generation
• Active Domain Knowledge Acquisition with 100-Dollar Budget: Enhancing LLMs via Cost-Efficient, Expert-Involved Interaction in Sensitive Domains
• Graph-R1: Incentivizing the Zero-Shot Graph Learning Capability in LLMs via Explicit Reasoning
• Natural Language Satisfiability: Exploring the Problem Distribution and Evaluating Transformer-based Language Models
• ActiveVLN: Towards Active Exploration via Multi-Turn RL in Vision-and-Language Navigation
• Selective Risk Certification for LLM Outputs via Information-Lift Statistics: PAC-Bayes, Robustness, and Skeleton Design
• LLM Hallucination Detection: A Fast Fourier Transform Method Based on Hidden Layer Temporal Signals
• Character-Centric Understanding of Animated Movies
• CBP-Tuning: Efficient Local Customization for Black-box Large Language Models
• A New Benchmark for Evaluating Code Translation with Third-Party Libraries
• AMQ: Enabling AutoML for Mixed-precision Weight-Only Quantization of Large Language Models
• Neuromorphic Intelligence
• Dr.V: A Hierarchical Perception-Temporal-Cognition Framework to Diagnose Video Hallucination by Fine-grained Spatial-Temporal Grounding
• Bhaasha, Bhasa, Zaban: A Survey for Low-Resourced Languages in South Asia -- Current Stage and Challenges
• D²HScore: Reasoning-Aware Hallucination Detection via Semantic Breadth and Depth Analysis in LLMs
• Phi: Preference Hijacking in Multi-modal Large Language Models at Inference Time
• Context-Aware Language Models for Forecasting Market Impact from Sequences of Financial News
• ToolRM: Outcome Reward Models for Tool-Calling Large Language Models
• Optimal Brain Restoration for Joint Quantization and Sparsification of LLMs
• ENJ: Optimizing Noise with Genetic Algorithms to Jailbreak LSMs
• Continually Adding New Languages to Multilingual Language Models
• Enhancing Generalization in Vision-Language-Action Models by Preserving Pretrained Representations
• Beyond Autoregression: An Empirical Study of Diffusion Large Language Models for Code Generation
• AI-Generated Content in Cross-Domain Applications: Research Trends, Challenges and Propositions
• HalluField: Detecting LLM Hallucinations via Field-Theoretic Modeling
• A Survey on Retrieval And Structuring Augmented Generation with Large Language Models
• DeepDive: Advancing Deep Search Agents with Knowledge Graphs and Multi-Turn RL
• RefactorCoderQA: Benchmarking LLMs for Multi-Domain Coding Question Solutions in Cloud and Edge Deployment
• Decoding Alignment: A Critical Survey of LLM Development Initiatives through Value-setting and Data-centric Lens
• URL2Graph++: Unified Semantic-Structural-Character Learning for Malicious URL Detection
• Explaining Black-box Language Models with Knowledge Probing Systems: A Post-hoc Explanation Perspective
• Multi-Intent Recognition in Dialogue Understanding: A Comparison Between Smaller Open-Source LLMs
• Unsupervised Hallucination Detection by Inspecting Reasoning Processes
• MCBP: A Memory-Compute Efficient LLM Inference Accelerator Leveraging Bit-Slice-enabled Sparsity and Repetitiveness
• Towards Understanding Visual Grounding in Visual Language Models
• ButterflyQuant: Ultra-low-bit LLM Quantization through Learnable Orthogonal Butterfly Transforms
• SimpleVLA-RL: Scaling VLA Training via Reinforcement Learning
• Boosting Embodied AI Agents through Perception-Generation Disaggregation and Asynchronous Pipeline Execution
• Combating the Memory Walls: Optimization Pathways for Long-Context Agentic LLM Inference
• ENSI: Efficient Non-Interactive Secure Inference for Large Language Models
• VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Model
• CESRec: Constructing Pseudo Interactions for Sequential Recommendation via Conversational Feedback
• LightAgent: Production-level Open-source Agentic AI Framework
• CCF: A Context Compression Framework for Efficient Long-Sequence Language Modeling
• HEFT: A Coarse-to-Fine Hierarchy for Enhancing the Efficiency and Accuracy of Language Model Reasoning
• Unbiased Reasoning for Knowledge-Intensive Tasks in Large Language Models via Conditional Front-Door Adjustment
• Compass-v3: Scaling Domain-Specific LLMs for Multilingual E-Commerce in Southeast Asia
• Towards Confidential and Efficient LLM Inference with Dual Privacy Protection
• CoDiCodec: Unifying Continuous and Discrete Compressed Representations of Audio
• When FinTech Meets Privacy: Securing Financial LLMs with Differential Private Fine-Tuning
• Toward a Multi-Echelon Cyber Warfare Theory: A Meta-Game-Theoretic Paradigm for Defense and Dominance
• Building High-Quality Datasets for Portuguese LLMs: From Common Crawl Snapshots to Industrial-Grade Corpora
• BitROM: Weight Reload-Free CiROM Architecture Towards Billion-Parameter 1.58-bit LLM Inference
• LLM-VeriPPA: Power, Performance, and Area Optimization aware Verilog Code Generation with Large Language Models
• RoboChemist: Long-Horizon and Safety-Compliant Robotic Chemical Experimentation
• Two Facets of the Same Optimization Coin: Model Degradation and Representation Collapse in Graph Foundation Models
• MERLIN: Multi-Stage Curriculum Alignment for Multilingual Encoder-LLM Integration in Cross-Lingual Reasoning
• Visual Representation Alignment for Multimodal Large Language Models
• GENUINE: Graph Enhanced Multi-level Uncertainty Estimation for Large Language Models
• Point Linguist Model: Segment Any Object via Bridged Large 3D-Language Model
• Query Expansion in the Age of Pre-trained and Large Language Models: A Comprehensive Survey
• Getting In Contract with Large Language Models -- An Agency Theory Perspective On Large Language Model Alignment
• Theoretical Analysis on how Learning Rate Warmup Accelerates Convergence
• TaDiCodec: Text-aware Diffusion Speech Tokenizer for Speech Language Modeling
• Interpreting the Effects of Quantization on LLMs
• XSRD-Net: EXplainable Stroke Relapse Detection
• Guarding Your Conversations: Privacy Gatekeepers for Secure Interactions with Cloud-Based AI Models
• Paladin: Defending LLM-enabled Phishing Emails with a New Trigger-Tag Paradigm
• MoGU V2: Toward a Higher Pareto Frontier Between Model Usability and Security
• HAVE: Head-Adaptive Gating and ValuE Calibration for Hallucination Mitigation in Large Language Models
• O³Afford: One-Shot 3D Object-to-Object Affordance Grounding for Generalizable Robotic Manipulation
• LoaQ: Layer-wise Output Approximation Quantization
• Text-Trained LLMs Can Zero-Shot Extrapolate PDE Dynamics, Revealing a Three-Stage In-Context Learning Mechanism
• Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraint
• Rethinking LLM Parametric Knowledge as Post-retrieval Confidence for Dynamic Retrieval and Reranking
• LatinX: Aligning a Multilingual TTS Model with Direct Preference Optimization
• Icon²: Aligning Large Language Models Using Self-Synthetic Preference Data via Inherent Regulation
• A Survey of the State-of-the-Art in Conversational Question Answering Systems
• Mitigating Spurious Correlations Between Question and Answer via Chain-of-Thought Correctness Perception Distillation
• SpecPrune-VLA: Accelerating Vision-Language-Action Models via Action-Aware Self-Speculative Pruning
• LatticeWorld: A Multimodal Large Language Model-Empowered Framework for Interactive Complex World Generation
• HoPE: Hyperbolic Rotary Positional Encoding for Stable Long-Range Dependency Modeling in Large Language Models
• PropVG: End-to-End Proposal-Driven Visual Grounding with Multi-Granularity Discrimination
• A Study of Large Language Models for Patient Information Extraction: Model Architecture, Fine-Tuning Strategy, and Multi-task Instruction Tuning
• CTCC: A Robust and Stealthy Fingerprinting Framework for Large Language Models via Cross-Turn Contextual Correlation Backdoor
• SpikingBrain: Spiking Brain-inspired Large Models
• Polysemantic Dropout: Conformal OOD Detection for Specialized LLMs
• Communication-Efficient Collaborative LLM Inference via Distributed Speculative Decoding
• Towards a Unified View of Large Language Model Post-Training
• How Small is Enough? Empirical Evidence of Quantized Small Language Models for Automated Program Repair
• HAMSA: Hijacking Aligned Compact Models via Stealthy Automation
• FPC-VLA: A Vision-Language-Action Framework with a Supervisor for Failure Prediction and Correction
• Skywork UniPic 2.0: Building Kontext Model with Online RL for Unified Multimodal Model
• IPA: An Information-Reconstructive Input Projection Framework for Efficient Foundation Model Adaptation
• Differentiable Entropy Regularization: A Complexity-Aware Approach for Neural Optimization
• Cetvel: A Unified Benchmark for Evaluating Language Understanding, Generation and Cultural Capacity of LLMs for Turkish
• An Empirical Analysis of Discrete Unit Representations in Speech Language Modeling Pre-training
• TeRA: Vector-based Random Tensor Network for High-Rank Adaptation of Large Language Models
• RecBase: Generative Foundation Model Pretraining for Zero-Shot Recommendation
• EverTracer: Hunting Stolen Large Language Models via Stealthy and Robust Probabilistic Fingerprint
• Mixture-of-Clustered-Experts: Advancing Expert Specialization and Generalization in Instruction Tuning
• OPRA-Vis: Visual Analytics System to Assist Organization-Public Relationship Assessment with Large Language Models
• LLMCDSR: Enhancing Cross-Domain Sequential Recommendation with Large Language Models
• MLP-Offload: Multi-Level, Multi-Path Offloading for LLM Pre-training to Break the GPU Memory Wall
• Generative AI for Crystal Structures: A Review
• Implicit Reasoning in Large Language Models: A Comprehensive Survey
• An Epidemiological Knowledge Graph extracted from the World Health Organization's Disease Outbreak News
• How Instruction-Tuning Imparts Length Control: A Cross-Lingual Mechanistic Analysis
• Behavioral Fingerprinting of Large Language Models
• Preserving Bilinear Weight Spectra with a Signed and Shrunk Quadratic Activation Function
• Extracting OPQRST in Electronic Health Records using Large Language Models with Reasoning
• The Fools are Certain; the Wise are Doubtful: Exploring LLM Confidence in Code Completion
• Spacetime-GR: A Spacetime-Aware Generative Model for Large Scale Online POI Recommendation
• Text Takes Over: A Study of Modality Bias in Multimodal Intent Detection
• LobRA: Multi-tenant Fine-tuning over Heterogeneous Data
• REFRAG: Rethinking RAG based Decoding
• Generative Foundation Model for Structured and Unstructured Electronic Health Records
• Enhancing Large Language Model for Knowledge Graph Completion via Structure-Aware Alignment-Tuning
• Identifying Origins of Place Names via Retrieval Augmented Generation
• Supervised In-Context Fine-Tuning for Generative Sequence Labeling
• Exploring and Mitigating Fawning Hallucinations in Large Language Models
• Unlocking the Effectiveness of LoRA-FP for Seamless Transfer Implantation of Fingerprints in Downstream Models
• Any-Order Flexible Length Masked Diffusion
• DTRNet: Dynamic Token Routing Network to Reduce Quadratic Costs in Transformers
• PREE: Towards Harmless and Adaptive Fingerprint Editing in Large Language Models via Knowledge Prefix Enhancement
• Make me an Expert: Distilling from Generalist Black-Box Models into Specialized Models for Semantic Segmentation
• Two Causes, Not One: Rethinking Omission and Fabrication Hallucinations in MLLMs
• SurgLLM: A Versatile Large Multimodal Model with Spatial Focus and Temporal Awareness for Surgical Video Understanding
• GraphKV: Breaking the Static Selection Paradigm with Graph-Based KV Cache Eviction
• Balanced Actor Initialization: Stable RLHF Training of Distillation-Based Reasoning Models
• What Are Research Hypotheses?
• Odyssey: Adaptive Policy Selection for Resilient Distributed Training
• Evaluating Recabilities of Foundation Models: A Multi-Domain, Multi-Dataset Benchmark
• MM-SeR: Multimodal Self-Refinement for Lightweight Image Captioning
• Automatic Reviewers Fail to Detect Faulty Reasoning in Research Papers: A New Counterfactual Evaluation Framework
• PDTrim: Targeted Pruning for Prefill-Decode Disaggregation in Inference
• Improving Fisher Information Estimation and Efficiency for LoRA-based LLM Unlearning
• Decoding Memories: An Efficient Pipeline for Self-Consistency Hallucination Detection
• PromptSleuth: Detecting Prompt Injection via Semantic Intent Invariance
• MSRS: Evaluating Multi-Source Retrieval-Augmented Generation
• Lethe: Purifying Backdoored Large Language Models with Knowledge Dilution
• Improving Alignment in LVLMs with Debiased Self-Judgment
• OneRec-V2 Technical Report
• GUARD: Glocal Uncertainty-Aware Robust Decoding for Effective and Efficient Open-Ended Text Generation
• SUMMA: A Multimodal Large Language Model for Advertisement Summarization
• OLMoASR: Open Models and Data for Training Robust Speech Recognition Models
• Discrete Diffusion VLA: Bringing Discrete Diffusion to Action Decoding in Vision-Language-Action Policies
• Model Science: getting serious about verification, explanation and control of AI systems
• Pruning Weights but Not Truth: Safeguarding Truthfulness While Pruning LLMs
• SoK: Large Language Model Copyright Auditing via Fingerprinting
• TComQA: Extracting Temporal Commonsense from Text
• PSO-Merging: Merging Models Based on Particle Swarm Optimization
• Spotlight Attention: Towards Efficient LLM Generation via Non-linear Hashing-based KV Cache Retrieval
• Charting the Future of Scholarly Knowledge with AI: A Community Perspective
• CrystalICL: Enabling In-Context Learning for Crystal Generation
• LFD: Layer Fused Decoding to Exploit External Knowledge in Retrieval-Augmented Generation
• IELDG: Suppressing Domain-Specific Noise with Inverse Evolution Layers for Domain Generalized Semantic Segmentation
• Bi-LoRA: Efficient Sharpness-Aware Minimization for Fine-Tuning Large-Scale Models
• Alignment with Fill-In-the-Middle for Enhancing Code Generation
• MQAD: A Large-Scale Question Answering Dataset for Training Music Large Language Models
• Towards 6G Intelligence: The Role of Generative AI in Future Wireless Networks
• How Multimodal LLMs Solve Image Tasks: A Lens on Visual Grounding, Task Reasoning, and Answer Decoding
• SynthCoder: A Synthetical Strategy to Tune LLMs for Code Completion
• Even Heads Fix Odd Errors: Mechanistic Discovery and Surgical Repair in Transformer Attention
• One Joke to Rule them All? On the (Im)possibility of Generalizing Humor
• MemoryVLA: Perceptual-Cognitive Memory in Vision-Language-Action Models for Robotic Manipulation
• StepWiser: Stepwise Generative Judges for Wiser Reasoning
• ZeST: an LLM-based Zero-Shot Traversability Navigation for Unknown Environments
• An Investigation on Group Query Hallucination Attacks
• ClusterFusion: Expanding Operator Fusion Scope for LLM Inference via Cluster-Level Collective Primitive
• UniC-RAG: Universal Knowledge Corruption Attacks to Retrieval-Augmented Generation
• UrgenGo: Urgency-Aware Transparent GPU Kernel Launching for Autonomous Driving
• Principled Detection of Hallucinations in Large Language Models via Multiple Testing
• Better Language Model-Based Judging Reward Modeling through Scaling Comprehension Boundaries
• SEAM: Semantically Equivalent Across Modalities Benchmark for Vision-Language Models
• Stand on The Shoulders of Giants: Building JailExpert from Previous Attack Experience
• VocabTailor: Dynamic Vocabulary Selection for Downstream Tasks in Small Language Models
• Self-Guided Function Calling in Large Language Models via Stepwise Experience Recall
• Multi-IaC-Eval: Benchmarking Cloud Infrastructure as Code Across Multiple Formats
• Select to Know: An Internal-External Knowledge Self-Selection Framework for Domain-Specific Question Answering
• SafeLLM: Unlearning Harmful Outputs from Large Language Models against Jailbreak Attacks
• Reliable Unlearning Harmful Information in LLMs with Metamorphosis Representation Projection
• Transduction is All You Need for Structured Data Workflows
• Fine-tuning and prompt engineering for large language models-based code review automation
• Unveiling Trust in Multimodal Large Language Models: Evaluation, Analysis, and Mitigation
• S3LoRA: Safe Spectral Sharpness-Guided Pruning in Adaptation of Agent Planner
• Quantization Meets dLLMs: A Systematic Study of Post-training Quantization for Diffusion LLMs
• Universal and Transferable Adversarial Attack on Large Language Models Using Exponentiated Gradient Descent
• H2EAL: Hybrid-Bonding Architecture with Hybrid Sparse Attention for Efficient Long-Context LLM Inference
• Evaluating Sparse Autoencoders for Monosemantic Representation
• Two Birds with One Stone: Multi-Task Detection and Attribution of LLM-Generated Text
• The Collaboration Paradox: Why Generative AI Requires Both Strategic Intelligence and Operational Stability in Supply Chain Management
• ComputerRL: Scaling End-to-End Online Reinforcement Learning for Computer Use Agents
• GLASS: Global-Local Aggregation for Inference-time Sparsification of LLMs
• Reinforced Context Order Recovery for Adaptive Reasoning and Planning
• ONG: One-Shot NMF-based Gradient Masking for Efficient Model Sparsification
• Hallucinations in medical devices
• Learning to Steer: Input-dependent Steering for Multimodal LLMs
• Involuntary Jailbreak: On Self-Prompting Attacks
• Prompt-Induced Linguistic Fingerprints for LLM-Generated Fake News Detection
• GTool: Graph Enhanced Tool Planning with Large Language Model
• Consiglieres in the Shadow: Understanding the Use of Uncensored Large Language Models in Cybercrimes
• Enhancing Cryptocurrency Sentiment Analysis with Multimodal Features
• Energy-Efficient Wireless LLM Inference via Uncertainty and Importance-Aware Speculative Decoding
• Can Large Models Teach Student Models to Solve Mathematical Problems Like Human Beings? A Reasoning Distillation Method via Multi-LoRA Interaction
• Is GPT-OSS Good? A Comprehensive Evaluation of OpenAI's Latest Open Source Models
• Uncovering Emergent Physics Representations Learned In-Context by Large Language Models
• ReaLM: Reflection-Enhanced Autonomous Reasoning with Small Language Models
• Rethinking Safety in LLM Fine-tuning: An Optimization Perspective
• AI-Augmented CI/CD Pipelines: From Code Commit to Production with Autonomous Decisions
• Dropping Just a Handful of Preferences Can Change Top Large Language Model Rankings
• STEM: Efficient Relative Capability Evaluation of LLMs through Structured Transition Samples
• Human Centric General Physical Intelligence for Agile Manufacturing Automation
• MoE-TTS: Enhancing Out-of-Domain Text Understanding for Description-based TTS via Mixture-of-Experts
• ADMIRE-BayesOpt: Accelerated Data MIxture RE-weighting for Language Models with Bayesian Optimization
• MM-BrowseComp: A Comprehensive Benchmark for Multimodal Browsing Agents
• MSRS: Adaptive Multi-Subspace Representation Steering for Attribute Alignment in Large Language Models
• Reverse Physician-AI Relationship: Full-process Clinical Diagnosis Driven by a Large Language Model
• Dataset Construction for Training LLM to Learn Analog Circuit Knowledge
• Layer-Wise Perturbations via Sparse Autoencoders for Adversarial Text Generation
• Large Model Empowered Embodied AI: A Survey on Decision-Making and Embodied Learning
• XQuant: Breaking the Memory Wall for LLM Inference with KV Cache Rematerialization
• Improving Generative Cross-lingual Aspect-Based Sentiment Analysis with Constrained Decoding
• Meta-Metrics and Best Practices for System-Level Inference Performance Benchmarking
• Chain-of-Query: Unleashing the Power of LLMs in SQL-Aided Table Understanding via Multi-Agent Collaboration
• MEML-GRPO: Heterogeneous Multi-Expert Mutual Learning for RLVR Advancement
• On Negative-aware Preference Optimization for Recommendation
• Preacher: Paper-to-Video Agentic System
• The Surprising Effectiveness of Membership Inference with Simple N-Gram Coverage
• A Chain of Diagnosis Framework for Accurate and Explainable Radiology Report Generation
• MiCo: End-to-End Mixed Precision Neural Network Co-Exploration Framework for Edge AI
• NeuronTune: Fine-Grained Neuron Modulation for Balanced Safety-Utility Alignment in LLMs
• EGGS-PTP: An Expander-Graph Guided Structured Post-training Pruning Method for Large Language Models
• Faster and Memory-Efficient Training of Sequential Recommendation Models for Large Catalogs
• A Signer-Invariant Conformer and Multi-Scale Fusion Transformer for Continuous Sign Language Recognition
• SMA: Who Said That? Auditing Membership Leakage in Semi-Black-box RAG Controlling
• Scaling Up Active Testing to Large Language Models
• READER: Retrieval-Assisted Drafter for Efficient LLM Inference
• Oblivionis: A Lightweight Learning and Unlearning Framework for Federated Large Language Models
• EditMF: Drawing an Invisible Fingerprint for Your Large Language Models
• Prompt-and-Check: Using Large Language Models to Evaluate Communication Protocol Compliance in Simulation-Based Training
• UWB at WASSA-2024 Shared Task 2: Cross-lingual Emotion Detection
• KFFocus: Highlighting Keyframes for Enhanced Video Understanding
• Large Language Models in the Data Science Lifecycle: A Systematic Mapping Study
• ExpVG: Investigating the Design Space of Visual Grounding in Multimodal Large Language Model
• DiffractGPT: Atomic Structure Determination from X-ray Diffraction Patterns using Generative Pre-trained Transformer
• Being-M0.5: A Real-Time Controllable Vision-Language-Motion Model
• A Data-driven ML Approach for Maximizing Performance in LLM-Adapter Serving
• Auditory Intelligence: Understanding the World Through Sound
• Towards Multimodal Sentiment Analysis via Contrastive Cross-modal Retrieval Augmentation and Hierachical Prompts
• MAViS: A Multi-Agent Framework for Long-Sequence Video Storytelling
• Large Language Models for Subjective Language Understanding: A Survey
• Large Language Models for Czech Aspect-Based Sentiment Analysis
• SynLLM: A Comparative Analysis of Large Language Models for Medical Tabular Synthetic Data Generation via Prompt Engineering
• VISOR: Visual Input-based Steering for Output Redirection in Vision-Language Models
• Pareto Multi-Objective Alignment for Language Models
• AimBot: A Simple Auxiliary Visual Cue to Enhance Spatial Awareness of Visuomotor Policies
• Modeling and Detecting Company Risks from News: A Case Study in Bloomberg News
• HSA-Net: Hierarchical and Structure-Aware Framework for Efficient and Scalable Molecular Language Modeling
• Large-scale Multi-sequence Pretraining for Generalizable MRI Analysis in Versatile Clinical Applications
• Enhancing Rumor Detection Methods with Propagation Structure Infused Language Model
• LP-Spec: Leveraging LPDDR PIM for Efficient LLM Mobile Speculative Inference with Architecture-Dataflow Co-Optimization
• HealthBranches: Synthesizing Clinically-Grounded Question Answering Datasets via Decision Pathways
• Tasa: Thermal-aware 3D-Stacked Architecture Design with Bandwidth Sharing for LLM Inference
• Many-Turn Jailbreaking
• Position: Ideas Should be the Center of Machine Learning Research
• Multimodal learning with next-token prediction for large multimodal models
• Generative Artificial Intelligence Extracts Structure-Function Relationships from Plants for New Materials
• When a Paper Has 1000 Authors: Rethinking Citation Metrics in the Era of LLMs
• PASG: A Closed-Loop Framework for Automated Geometric Primitive Extraction and Semantic Anchoring in Robotic Manipulation
• ASAudio: A Survey of Advanced Spatial Audio Research
• NEP: Autoregressive Image Editing via Next Editing Token Prediction
• Comparing Knowledge Injection Methods for LLMs in a Low-Resource Regime
• Diffusion LLMs Can Do Faster-Than-AR Inference via Discrete Diffusion Forcing
• Learning by Teaching: Engaging Students as Instructors of Large Language Models in Computer Science Education
• Omni Geometry Representation Learning vs Large Language Models for Geospatial Entity Resolution
• LLMEval-Fair: A Large-Scale Longitudinal Study on Robust and Fair Evaluation of Large Language Models
• UNCAGE: Contrastive Attention Guidance for Masked Generative Transformers in Text-to-Image Generation
• CT-GRAPH: Hierarchical Graph Attention Network for Anatomy-Guided CT Report Generation
• Simulating LLM training workloads for heterogeneous compute and network infrastructure
• Decision-Making with Deliberation: Meta-reviewing as a Document-grounded Dialogue
• SPEX: A Vision-Language Model for Land Cover Extraction on Spectral Remote Sensing Images
• Attention Basin: Why Contextual Position Matters in Large Language Models
• Align, Don't Divide: Revisiting the LoRA Architecture in Multi-Task Learning
• Towards Hallucination-Free Music: A Reinforcement Learning Preference Optimization Framework for Reliable Song Generation
• Tesserae: Scalable Placement Policies for Deep Learning Workloads
• iFairy: the First 2-bit Complex LLM with All Parameters in \±1, ± i\
• Adversarial Attacks and Defenses on Graph-aware Large Language Models (LLMs)
• Enhancing Dialogue Annotation with Speaker Characteristics Leveraging a Frozen LLM
• SEAgent: Self-Evolving Computer Use Agent with Autonomous Learning from Experience
• FaST: Feature-aware Sampling and Tuning for Personalized Preference Alignment with Limited Data
• Open Scene Graphs for Open-World Object-Goal Navigation
• GeRe: Towards Efficient Anti-Forgetting in Continual Learning of LLM via General Samples Replay
• X-SAM: From Segment Anything to Any Segmentation
• A Reproducible, Scalable Pipeline for Synthesizing Autoregressive Model Literature
• GraphProp: Training the Graph Foundation Models using Graph Properties
• CARD: A Cache-Assisted Parallel Speculative Decoding Framework via Query-and-Correct Paradigm for Accelerating LLM Inference
• Boosting Visual Knowledge-Intensive Training for LVLMs Through Causality-Driven Visual Object Completion
• FlexQ: Efficient Post-training INT6 Quantization for LLM Serving via Algorithm-System Co-Design
• Forgetting: A New Mechanism Towards Better Large Language Model Fine-tuning
• Large Language Model's Multi-Capability Alignment in Biomedical Domain
• Audio Does Matter: Importance-Aware Multi-Granularity Fusion for Video Moment Retrieval
• Leveraging large language models for SQL behavior-based database intrusion detection
• Hierarchical Text Classification Using Black Box Large Language Models
• WARP -- Web-Augmented Real-time Program Repairer: A Real-Time Compilation Error Resolution using LLMs and Web-Augmented Synthesis
• Gather and Trace: Rethinking Video TextVQA from an Instance-oriented Perspective
• Efficient Scaling for LLM-based ASR
• DTPA: Dynamic Token-level Prefix Augmentation for Controllable Text Generation
• Majority Bit-Aware Watermarking For Large Language Models
• Putnam-AXIOM: A Functional and Static Benchmark for Measuring Higher Level Mathematical Reasoning in LLMs
• LLMDistill4Ads: Using Cross-Encoders to Distill from LLM Signals for Advertiser Keyphrase Recommendations
• Draw Your Mind: Personalized Generation via Condition-Level Modeling in Text-to-Image Diffusion Models
• R2GenKG: Hierarchical Multi-modal Knowledge Graph for LLM-based Radiology Report Generation
• A Comparative Study of Neurosymbolic AI Approaches to Interpretable Logical Reasoning
• VLMQ: Efficient Post-Training Quantization for Large Vision-Language Models via Hessian Augmentation
• Exploring Layer-wise Information Effectiveness for Post-Training Quantization in Small Language Models
• MedBLINK: Probing Basic Perception in Multimodal Language Models for Medicine
• GrandJury: A Collaborative Machine Learning Model Evaluation Protocol for Dynamic Quality Rubrics
• A Multi-Agent System for Complex Reasoning in Radiology Visual Question Answering
• Defending Against Knowledge Poisoning Attacks During Retrieval-Augmented Generation
• ReGate: Enabling Power Gating in Neural Processing Units
• LaMPE: Length-aware Multi-grained Positional Encoding for Adaptive Long-context Scaling Without Training
• Kron-LoRA: Hybrid Kronecker-LoRA Adapters for Scalable, Sustainable Fine-tuning
• Charting 15 years of progress in deep learning for speech emotion recognition: A replication study
• Prefill-Decode Aggregation or Disaggregation? Unifying Both for Goodput-Optimized LLM Serving
• Isolating Culture Neurons in Multilingual Large Language Models
• Alleviating Attention Hacking in Discriminative Reward Modeling through Interaction Distillation
• PoeTone: A Framework for Constrained Generation of Structured Chinese Songci with LLMs
• GlaBoost: A multimodal Structured Framework for Glaucoma Risk Stratification
• MLP Memory: A Retriever-Pretrained Memory for Large Language Models
• Intention-Guided Cognitive Reasoning for Egocentric Long-Term Action Anticipation
• Enhancing the Preference Extractor in Multi-turn Dialogues: From Annotating Disasters to Accurate Preference Extraction
• EAC-MoE: Expert-Selection Aware Compressor for Mixture-of-Experts Large Language Models
• MagicVL-2B: Empowering Vision-Language Models on Mobile Devices with Lightweight Visual Encoders via Curriculum Learning
• JSidentify-V2: Leveraging Dynamic Memory Fingerprinting for Mini-Game Plagiarism Detection
• Aligning Language Models with Real-time Knowledge Editing
• Multi-TW: Benchmarking Multimodal Models on Traditional Chinese Question Answering in Taiwan
• MeshLLM: Empowering Large Language Models to Progressively Understand and Generate 3D Mesh
• A Dynamic Allocation Scheme for Adaptive Shared-Memory Mapping on Kilo-core RV Clusters for Attention-Based Model Deployment
• Effective Damage Data Generation by Fusing Imagery with Human Knowledge Using Vision-Language Models
• CSIRO-LT at SemEval-2025 Task 11: Adapting LLMs for Emotion Recognition for Multiple Languages
• DisTaC: Conditioning Task Vectors via Distillation for Robust Model Merging
• ROVI: A VLM-LLM Re-Captioned Dataset for Open-Vocabulary Instance-Grounded Text-to-Image Generation
• Beyond Fixed: Training-Free Variable-Length Denoising for Diffusion Large Language Models
• Latent Knowledge Scalpel: Precise and Massive Knowledge Editing for Large Language Models
• Instruction-Grounded Visual Projectors for Continual Learning of Generative Vision-Language Models
• From Individuals to Crowds: Dual-Level Public Response Prediction in Social Media
• CoT-Self-Instruct: Building high-quality synthetic prompts for reasoning and non-reasoning tasks
• Deep Learning-based Prediction of Clinical Trial Enrollment with Uncertainty Estimates
• Causal2Vec: Improving Decoder-only LLMs as Versatile Embedding Models
• Unveiling Super Experts in Mixture-of-Experts Large Language Models
• PixNerd: Pixel Neural Field Diffusion
• Fine-Grained Privacy Extraction from Retrieval-Augmented Generation Systems via Knowledge Asymmetry Exploitation
• Not Just What, But When: Integrating Irregular Intervals to LLM for Sequential Recommendation
• LENS: Learning Ensemble Confidence from Neural States for Multi-LLM Answer Integration
• Phi-Ground Tech Report: Advancing Perception in GUI Grounding
• Counterfactual Evaluation for Blind Attack Detection in LLM-based Evaluation Systems
• Learning Like Humans: Resource-Efficient Federated Fine-Tuning through Cognitive Developmental Stages
• Distilling Knowledge from Large Language Models: A Concept Bottleneck Model for Hate and Counter Speech Recognition
• KLLM: Fast LLM Inference with K-Means Quantization
• LOTS of Fashion! Multi-Conditioning for Image Generation via Sketch-Text Pairing
• Exploiting Synergistic Cognitive Biases to Bypass Safety in LLMs
• Beyond Single Labels: Improving Conversational Recommendation through LLM-Powered Data Augmentation
• IFEvalCode: Controlled Code Generation
• What is an "Abstract Reasoner"? Revisiting Experiments and Arguments about Large Language Models

Related

• LLaMA-Omni2: LLM-based Real-time Spoken Chatbot with Autoregressive Streaming Speech Synthesis
• Large Language Models: A Survey
• Three Ways of Using Large Language Models to Evaluate Chat
• The Llama 3 Herd of Models
• YAYI 2: Multilingual Open-Source Large Language Models