AI Paper Atlas

Research map / Benchmarks / OlympiadBench

OlympiadBench leaderboard

62 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on OlympiadBench. Best accuracy: 75.1 (Prompt-Hint-SFT-32B, NeurIPS 2025).

Numbers are as each paper reports them, read by AI from its results tables and not checked by hand. Papers can differ in setting (training data, model size, extra data); the topic column shows what each paper is about. Open the paper before you compare or quote a number.

OlympiadBench · accuracy

35 papers, each with its best reported number for its own method; higher is better.

#accuracy ↑MethodPaperTopic
175.1Prompt-Hint-SFT-32B
DeepSeek-R1-32B
Teaching Language Models to Reason with Tools
NeurIPS 2025 · Table 1
Reasoning and chain of thought
269.02DAPO w/ Forking Tokens
Qwen3-32B
Beyond the 80/20 Rule: High-Entropy Minority Tokens Drive Effective Reinforcement Learning for LLM Reasoning
NeurIPS 2025 · Table: comparison between vani
Reasoning and chain of thought
365.88DAPO w/ EDAS
Qwen3-8B
Leveraging Error Diversity in Group Rollouts for Reinforcement Learning
NeurIPS 2026 · Table 1
RL from feedback
464.11R1-Distill-Qwen-32B + Steering
R1-Distill-Qwen-32B
On Reasoning Strength Planning in Large Reasoning Models
NeurIPS 2025 · Table: Accuracy comparison bef
Reasoning and chain of thought
558.6Spectral-guided Learning
Qwen3-4B-Instruct-2507
Uncovering the Gradient Geometry of Long CoT: A Spectral-guided Approach to Reasoning Distillation
ICML 2026 · Table 1
Reasoning and chain of thought
657.6TRAPO
Qwen2.5-Math-7B
Trust-Region Adaptive Policy Optimization
ICLR 2026 · Table 1
RL from feedback
756.9ARES-RL-7B
Qwen2.5-VL-7B-Instruct
ARES: Multimodal Adaptive Reasoning via Difficulty-Aware Token-Level Entropy Shaping
ICLR 2026 · Table 4
Reasoning and chain of thought
853.31SBSC (greedy)
Claude-3.5-Sonnet
SBSC: Step-by-Step Coding for Improving Mathematical Olympiad Performance
ICLR 2025 · Table 1
Reasoning and chain of thought
950.96Self-Verification (R1-1.5B)@16
Self-Verification-R1-1.5B
Incentivizing LLMs to Self-Verify Their Answers
NeurIPS 2025 · Table 3
Reasoning and chain of thought
1050.2VeriFree
Qwen3-8B
Reinforcing General Reasoning Without Verifiers
ICLR 2026 · Table 3
Reasoning and chain of thought
25 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

OlympiadBench · pass@1

27 papers, each with its best reported number for its own method; higher is better.

#pass@1 ↑MethodPaperTopic
170.7PADD
Qwen3-30B-A3B student
PADD: Path-Aligned Decompression Distillation for Non-Router Teacher to Guide MoE Student Learning
ICML 2026 · Table 1
Reasoning and chain of thought
268.8MathSmith-HC
Qwen3-8B (long-CoT)
MathSmith: Towards Extremely Hard Mathematical Reasoning by Forging Synthetic Problems with a Reinforced Policy
AAAI 2026 · Table 1
Reasoning and chain of thought
365.6rStar-Math (7B SLM+7B PPM)⁶⁴
Qwen2.5-Math-7B
rStar-Math: Small LLMs Can Master Math Reasoning with Self-Evolved Deep Thinking
ICML 2025 · Table: results on challenging
Reasoning and chain of thought
460.5SwS-32B
Qwen2.5-32B
SwS: Self-aware Weakness-driven Problem Synthesis in Reinforcement Learning for LLM Reasoning
NeurIPS 2025 · Table 1
Reasoning and chain of thought
557.2LUFFY
Qwen2.5-Math-7B
Learning to Reason under Off-Policy Guidance
NeurIPS 2025 · Table 1
RL from feedback
657.03Dr. MAMR
Qwen2.5-14B-Instruct
Unlocking the Power of Multi-Agent LLM for Reasoning: From Lazy Agents to Deliberation
ICLR 2026 · Table 1
LLM agents
755.3VSRM-PPO
DS-Distill-7B
Promoting Efficient Reasoning with Verifiable Stepwise Reward
AAAI 2026 · Table 1
Reasoning and chain of thought
854.8DS-R1-Distill-Qwen-7B + GAR
DeepSeek-R1-Distill-Qwen-7B
Generative Adversarial Reasoner: Enhancing LLM Reasoning with Adversarial Reinforcement Learning
ICLR 2026 · Table 1
Reasoning and chain of thought
954.5Step-RS2
DeepSeek-R1-Distill-Qwen-1.5B
Step-GRPO: Enhancing Reasoning Quality and Efficiency via Structured PRM-Based Reinforcement Learning
AAAI 2026 · Table 1
Reasoning and chain of thought
1053.3DeepCompress-Zero-7B
Qwen2.5-7B
DeepCompress: A Dual Reward Strategy for Dynamically Exploring and Compressing Reasoning Chains
ICLR 2026 · Table 1
Reasoning and chain of thought
17 more papers on this leaderboard, the baselines reported on it and a CSV download are free with a scime account.
Sign in free to see all

Not affiliated with any conference, journal or the benchmark's authors. Corrections: contact@latenticlabs.com.