BloombergGPT: A Large Language Model for Finance
2023/03/30 by Shijie Wu, Ozan Irsoy, Ozan İrsoy +16 · 2 voices · 134 citations
Computer Science · Decision Sciences · #Topic Modeling #Stock Market Forecasting Methods #Natural Language Processing Techniques
paper · pdf · doi:10.48550/arxiv.2303.17564
Abstract
The use of NLP in the realm of financial technology is broad and complex, with applications ranging from sentiment analysis and named entity recognition to question answering. Large Language Models (LLMs) have been shown to be effective on a variety of tasks; however, no LLM specialized for the financial domain has been reported in literature. In this work, we present BloombergGPT, a 50 billion parameter language model that is trained on a wide range of financial data. We construct a 363 billion token dataset based on Bloomberg's extensive data sources, perhaps the largest domain-specific dataset yet, augmented with 345 billion tokens from general purpose datasets. We validate BloombergGPT on standard LLM benchmarks, open financial benchmarks, and a suite of internal benchmarks that most accurately reflect our intended usage. Our mixed dataset training leads to a model that outperforms existing models on financial tasks by significant margins without sacrificing performance on general LLM benchmarks. Additionally, we explain our modeling choices, training process, and evaluation methodology. We release Training Chronicles (Appendix C) detailing our experience in training BloombergGPT.
Cited by
- Capital Markets LLM Reliability Score (CM-LRS): From Plausible to Bankable
- TriAgent: Divergence-Aware Multi-Agent Committees for Cost-Efficient Financial Sentiment Analysis
- Frontier AI performance across the business disciplines: a case-grounded benchmark of knowledge work and analytical reasoning
- FinRAG-12B: A Production-Validated Recipe for Grounded Question Answering in Banking
- FinSAgent: Corpus-Aligned Multi-Agent RAG Framework for Evidence-Grounded SEC Filing Question Answering
- Planning with Transformers: Chain of Computation and Structured Context Windows
- Octopus v4: Graph of language models
- AI Trading: Evaluating Large Language Models for Technical Market Analysis
- FORCE-Bench: A Benchmark, Dataset, and Evaluation Harness for Agentic AI in Enterprise Finance
- FinBench: Time-Gated Calibration and Uncertainty Benchmarking for Agentic Financial Forecasting
- DFAH-Bench: Benchmarking Observable Agent Instability in Financial Decision-Making
- InfoFlood: Jailbreaking Large Language Models with Information Overload
- TAID: Temporally Adaptive Interpolated Distillation for Efficient Knowledge Transfer in Language Models
- Alpha-R1: Alpha Screening with LLM Reasoning via Reinforcement Learning
- Accounting Reasoning in Large Language Models: Concepts, Evaluation, and Empirical Analysis
- IKS-Instruct: A 24,000-Example Multilingual Dataset for Teaching Language Models Indian Knowledge Systems
- FinAbstain: Uncertainty-Calibrated Multimodal RAG for Selective Financial Forecasting
- Toward Automated Detection of Documentation Inconsistencies in Electronic Health Records
- Evaluating Small Language Models for Agentic On-Farm Decision Support Systems
- Understanding Structured Financial Data with LLMs: A Case Study on Fraud Detection
- Large and Small Model Collaboration for Air Interface
- MixtureKit: A General Framework for Composing, Training, and Visualizing Mixture-of-Experts Models
- The Data Efficiency Frontier of Financial Foundation Models: Scaling Laws from Continued Pretraining
- Market-Bench: Evaluating Large Language Models on Introductory Quantitative Trading and Market Dynamics
- PyFi: Toward Pyramid-like Financial Image Understanding for VLMs via Adversarial Agents
- Financial News Summarization: Can extractive methods still offer a true alternative to LLMs?
- Is GPT-OSS All You Need? Benchmarking Large Language Models for Financial Intelligence and the Surprising Efficiency Paradox
- Evaluating Hydro-Science and Engineering Knowledge of Large Language Models
- Detecting AI Hallucinations in Finance: An Information-Theoretic Method Cuts Hallucination Rate by 92%
- LLM CHESS: Benchmarking Reasoning and Instruction-Following in LLMs through Chess
- LLM-Generated Counterfactual Stress Scenarios for Portfolio Risk Simulation via Hybrid Prompt-RAG Pipeline
- FISCAL: Financial Synthetic Claim-document Augmented Learning for Efficient Fact-Checking
- Diagram-to-Circuit QNLP for Financial Sentiment Analysis
- A symbolic Perl algorithm for the unification of Nahuatl word spellings
- Beyond GeneGPT: A Multi-Agent Architecture with Open-Source LLMs for Enhanced Genomic Question Answering
- Generalist Foundation Models Are Not Clinical Enough for Hospital Operations
- MM-Telco: Benchmarks and Multimodal Large Language Models for Telecom Applications
- Honesty over Accuracy: Trustworthy Language Models through Reinforced Hesitation
- LAET: A Layer-wise Adaptive Ensemble Tuning Framework for Pretrained Language Models
- History Rhymes: Macro-Contextual Retrieval for Robust Financial Forecasting
- Prompt Tuning for Natural Language to SQL with Embedding Fine-Tuning and RAG
- LoopLLM: Transferable Energy-Latency Attacks in LLMs via Repetitive Generation
- RedOne 2.0: Rethinking Domain-specific LLM Post-Training in Social Networking Services
- FinRpt: Dataset, Evaluation System and LLM-based Multi-agent Framework for Equity Research Report Generation
- FedRW: Efficient Privacy-Preserving Data Reweighting for Enhancing Federated Learning of Language Models
- Synthetic Data-Driven Prompt Tuning for Financial QA over Tables and Documents
- Reasoning on Time-Series for Financial Technical Analysis
- Design principles for text-to-image generative artificial intelligence creativity support tools for visual design
- Credit Network Modeling and Analysis via Large Language Models
- Merging Continual Pretraining Models for Domain-Specialized LLMs: A Case Study in Finance
- Evontree: Ontology Rule-Guided Self-Evolution of Large Language Models
- HACK: Hallucinations Along Certainty and Knowledge Axes
- P1GPT: a multi-agent LLM workflow module for multi-modal financial information analysis
- SEGA: A Stepwise Evolution Paradigm for Content-Aware Layout Generation with Design Prior
- News-Aware Direct Reinforcement Trading for Financial Markets
- Trading with the Devil: Risk and Return in Foundation Model Strategies
- Investigating the Impact of Rationales for LLMs on Natural Language Understanding
- Advances in Pre-trained Language Models for Domain-Specific Text Classification: A Systematic Review
- Structure-R1: Dynamically Leveraging Structural Knowledge in LLM Reasoning through Reinforcement Learning
- FinAI Data Assistant: LLM-based Financial Database Query Processing with the OpenAI Function Calling API
- Mirror Speculative Decoding: Breaking the Serial Barrier in LLM Inference
- Program of Thoughts for Financial Reasoning: Leveraging Dynamic In-Context Examples and Generative Retrieval
- Aligning Language Models with Investor and Market Behavior for Financial Recommendations
- StockBench: Can LLM Agents Trade Stocks Profitably In Real-world Markets?
- Integrating Large Language Models and Reinforcement Learning for Sentiment-Driven Quantitative Trading
- StelLA: Subspace Learning in Low-rank Adaptation using Stiefel Manifold
- Exploring Large Language Models for Financial Applications: Techniques, Performance, and Challenges with FinMA
- Attention Once Is All You Need: Efficient Streaming Inference with Stateful Transformers
- Adversarial News and Lost Profits: Manipulating Headlines in LLM-Driven Algorithmic Trading
- GuruAgents: Emulating Wise Investors with Prompt-Guided LLM Agents
- Profit Mirage: Revisiting Information Leakage in LLM-based Financial Agents
- An Adaptive Multi Agent Bitcoin Trading System
- DACIP-RC: Domain Adaptive Continual Instruction Pre-Training via Reading Comprehension on Business Conversations
- Impact of LLMs on Team Collaboration in Software Development
- DACP: Domain-Adaptive Continual Pre-Training of Large Language Models for Phone Conversation Summarization
- The New Quant: A Survey of Large Language Models in Financial Prediction and Trading
- QuantAgents: Towards Multi-agent Financial System via Simulated Trading
- Large Language Models Hallucination: A Comprehensive Survey
- SECA: Semantically Equivalent and Coherent Attacks for Eliciting LLM Hallucinations
- Can LLMs Hit Moving Targets? Tracking Evolving Signals in Corporate Disclosures
- One More Question is Enough, Expert Question Decomposition (EQD) Model for Domain Quantitative Reasoning
- Memory-Augmented Log Analysis with Phi-4-mini: Enhancing Threat Detection in Structured Security Logs
- Structuring Reasoning for Complex Rules Beyond Flat Representations
- Quantifying Semantic Shift in Financial NLP: Robust Metrics for Market Prediction Stability
- Finetune Once: Decoupling General & Domain Learning with Dynamic Boosted Annealing
- Better with Less: Small Proprietary Models Surpass Large Language Models in Financial Transaction Understanding
- Evolution Strategies at Scale: LLM Fine-Tuning Beyond Reinforcement Learning
- Comparing Open-Source and Commercial LLMs for Domain-Specific Analysis and Reporting: Software Engineering Challenges and Design Trade-offs
- Fin-Ally: Pioneering the Development of an Advanced, Commonsense-Embedded Conversational AI for Money Matters
- Generative AI Solutions to Empower Financial Firms
- RHYTHM: Reasoning with Hierarchical Temporal Tokenization for Human Mobility
- Evaluating Open-Source Large Language Models for Technical Telecom Question Answering
- You Can't Steal Nothing: Mitigating Prompt Leakages in LLMs via System Vectors
- KnowMT-Bench: Benchmarking Knowledge-Intensive Long-Form Question Answering in Multi-Turn Dialogues
- QuantMind: A Context-Engineering Based Knowledge Framework for Quantitative Finance
- Unlocking Financial Insights: An advanced Multimodal Summarization with Multimodal Output Framework for Financial Advisory Videos
- Can Federated Learning Safeguard Private Data in LLM Training? Vulnerabilities, Attacks, and Defense Evaluation
- Beyond Sentiment: Structured Information Extraction from Financial News
- Can Large Language Models Execute Parent Orders?
- A scoping review of ChatGPT research in accounting and finance
- AI for social science and social science of AI: A survey
- Financial Risk Relation Identification through Dual-view Adaptation
- Confidential LLM Inference: Performance and Cost Across CPU and GPU TEEs
- Can an Individual Manipulate the Collective Decisions of Multi-Agents?
- Time to Revist Exact Match
- Enhancing Financial RAG with Agentic AI and Multi-HyDE: A Novel Approach to Knowledge Retrieval and Hallucination Reduction
- SynBench: A Benchmark for Differentially Private Text Generation
- Multi-Model Synthetic Training for Mission-Critical Small Language Models
- Analogy-Driven Financial Chain-of-Thought (AD-FCoT): A Prompting Approach for Financial Sentiment Analysis
- FinGEAR: Financial Mapping-Guided Enhanced Answer Retrieval
- THEME: Enhancing Thematic Investing with Semantic Stock Representations and Temporal Dynamics
- Compass-v3: Scaling Domain-Specific LLMs for Multilingual E-Commerce in Southeast Asia
- Building High-Quality Datasets for Portuguese LLMs: From Common Crawl Snapshots to Industrial-Grade Corpora
- A Role-Aware Multi-Agent Framework for Financial Education Question Answering with LLMs
- Getting In Contract with Large Language Models -- An Agency Theory Perspective On Large Language Model Alignment
- Towards EnergyGPT: A Large Language Model Specialized for the Energy Sector
- Uncovering the Vulnerability of Large Language Models in the Financial Domain via Risk Concealment
- MM-ARC: Multimodal Adaptive Routing of Capital with Robustness-Audited Strategy Pools
- SelfAug: Mitigating Catastrophic Forgetting in Retrieval-Augmented Generation via Distribution Self-Alignment
- TULIP: Adapting Open-Source Large Language Models for Underrepresented Languages and Specialized Financial Tasks
- LLMs for LLMs: A Structured Prompting Methodology for Long Legal Documents
- Survey of Specialized Large Language Model
- LLMs and Agentic AI in Insurance Decision-Making: Opportunities and Challenges For Africa
- RPKT: Learning What You Don't -- Know Recursive Prerequisite Knowledge Tracing in Conversational AI Tutors for Personalized Learning
- Securing Educational LLMs: A Generalised Taxonomy of Attacks on LLMs and DREAD Risk Assessment
- Can Large Language Models Integrate Spatial Data? Empirical Insights into Reasoning Strengths and Computational Weaknesses
- SVGen: Interpretable Vector Graphics Generation with Large Language Models
- OS Agents: A Survey on MLLM-based Agents for General Computing Devices Use
- Tree-of-Reasoning: Towards Complex Medical Diagnosis via Multi-Agent Reasoning with Evidence Tree
- FinWorld: An All-in-One Open-Source Platform for End-to-End Financial AI Research and Deployment
- ProKG-Dial: Progressive Multi-Turn Dialogue Construction with Domain Knowledge Graphs
- FinKario: Event-Enhanced Automated Construction of Financial Knowledge Graph
- ContestTrade: A Multi-Agent Trading System Based on Internal Contest Mechanism
- List of large language models [wikipedia]
Discussions
Related