AI Paper Atlas

Research map / Benchmarks / AMC

AMC leaderboard

18 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on AMC. Best accuracy: 96.6 (LIMO-V2-P, 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.

AMC · accuracy

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

#accuracy ↑MethodPaperTopic
196.6LIMO-V2-P
Qwen2.5-32B-Instruct
LIMOPro: Reasoning Refinement for Efficient and Effective Test-time Scaling
NeurIPS 2025 · Table 2
Reasoning and chain of thought
291.6FAPO-32B
Qwen2.5-32B
FAPO: Flawed-Aware Policy Optimization for Efficient and Reliable Reasoning
ICLR 2026 · Table (additional results)
RL from feedback
381JET
DS-Distill-Qwen-7B
Your Models Have Thought Enough: Training Large Reasoning Models to Stop Overthinking
ICLR 2026 · Table 2
Reasoning and chain of thought
475.9ShorterBetter-7B (α=2)
DeepSeek-R1-Distill-Qwen-7B
ShorterBetter: Guiding Reasoning Models to Find Optimal Inference Length for Efficient Reasoning
NeurIPS 2025 · 7B results table
Reasoning and chain of thought
574.3SRFT
Qwen2.5-Math-7B
SRFT: A Single-Stage Method with Supervised and Reinforcement Fine-Tuning for Reasoning
ICLR 2026 · Table 2
Instruction and fine-tuning
672.8ERA
Qwen2.5-Math-7B
Entropy Regularizing Activation: Boosting Continuous Control, Large Language Models, and Image Classification with Activation as Entropy Constraints
ICLR 2026 · Table 2
RL from feedback
768.2Prefix-RFT
Qwen2.5-Math-7B
Blending Supervised and Reinforcement Fine-Tuning with Prefix Sampling
ICML 2026 · Table 1
RL from feedback
866.2TRAPO
Qwen2.5-Math-7B
Trust-Region Adaptive Policy Optimization
ICLR 2026 · Table 1
RL from feedback
966.2LUFFY†
Qwen2.5-Math-7B
Learning to Reason under Off-Policy Guidance
NeurIPS 2025 · Table 1
RL from feedback
1065.1ReLIFT
Qwen2.5-Math-7B
Learning What Reinforcement Learning Can't: Interleaved Online Fine-Tuning for Hardest Questions
ICLR 2026 · Table 1
Instruction and fine-tuning
8 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.