EXAONE 3.5: Series of Large Language Models for Real-world Use Cases

2024/12/06 by Soyoung An, Research, LG AI, Kyung‐Hoon Bae +62 · 15 citations
Business, Management and Accounting · Computer Science · Decision Sciences · #Business Process Modeling and Analysis #Computation and Language (cs.CL) #Data Quality and Management #FOS: Computer and information sciences #Recommender Systems and Techniques

paper · pdf · doi:10.48550/arxiv.2412.04862

openalex publication_date 2024/12/06 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

This technical report introduces the EXAONE 3.5 instruction-tuned language models, developed and released by LG AI Research. The EXAONE 3.5 language models are offered in three configurations: 32B, 7.8B, and 2.4B. These models feature several standout capabilities: 1) exceptional instruction following capabilities in real-world scenarios, achieving the highest scores across seven benchmarks, 2) outstanding long-context comprehension, attaining the top performance in four benchmarks, and 3) competitive results compared to state-of-the-art open models of similar sizes across nine general benchmarks. The EXAONE 3.5 language models are open to anyone for research purposes and can be downloaded from https://huggingface.co/LGAI-EXAONE. For commercial use, please reach out to the official contact point of LG AI Research: contact_u[email protected].

Cited by

• Soft Inductive Bias Approach via Explicit Reasoning Perspectives in Inappropriate Utterance Detection Using Large Language Models
• KoSimpleQA: A Korean Factuality Benchmark with an Analysis of Reasoning LLMs
• When to Ensemble: Identifying Token-Level Points for Stable and Fast LLM Ensembling
• Assessing Socio-Cultural Alignment and Technical Safety of Sovereign LLMs
• Reference-Free Rating of LLM Responses via Latent Information
• Responsible AI Technical Report
• Everyday Physics in Korean Contexts: A Culturally Grounded Physical Reasoning Benchmark
• From Ground Trust to Truth: Disparities in Offensive Language Judgments on Contemporary Korean Political Discourse
• KoBLEX: Open Legal Question Answering with Multi-hop Reasoning
• Can You Trick the Grader? Adversarial Persuasion of LLM Judges
• EXAONE 4.0: Unified Large Language Models Integrating Non-reasoning and Reasoning Modes
• Journalism-Guided Agentic In-Context Learning for News Stance Detection
• From KMMLU-Redux to KMMLU-Pro: A Professional Korean Benchmark Suite for LLM Evaluation
• PhoniTale: Phonologically Grounded Mnemonic Generation for Typologically Distant Language Pairs
• LLM-based Question-Answer Framework for Sensor-driven HVAC System Interaction

Related

• EXAONE 3.0 7.8B Instruction Tuned Language Model
• EXAONE 4.0: Unified Large Language Models Integrating Non-reasoning and Reasoning Modes
• EXAONE Deep: Reasoning Enhanced Language Models
• INSTRUCTEVAL: Towards Holistic Evaluation of Instruction-Tuned Large Language Models
• A Survey of Large Language Models