AI Paper Atlas

Research map / Benchmarks / GSM-Hard

GSM-Hard leaderboard

15 papers at ICML, NeurIPS, ICLR, CVPR, AAAI, ICCV and ECCV, and the journals TPAMI, IJCV and JMLR report results on GSM-Hard. Best accuracy: 76.1 (M-Pilot, 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.

GSM-Hard · accuracy

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

#accuracy ↑MethodPaperTopic
176.1M-Pilot
gpt-3.5-turbo
Matryoshka Pilot: Learning to Drive Black-Box LLMs with LLMs
NeurIPS 2025 · GSM8K results table
Language, safety and interpretability
273.5DEAR
Qwen3-8B
Group Perspective Matters: Regulating Debate Relationships Can Mitigate Blind Conformity in Multi-Agent Debate
NeurIPS 2026 · Appendix Table (Qwen3-8B)
LLM agents
373.32BRIDGE 2g
Gemini 1.5 Pro
From Few to Many: Self-Improving Many-Shot Reasoners Through Iterative Optimization and Generation
ICLR 2025 · Table: MATH and GSM-Hard
Language, safety and interpretability
466.6OpenMath-CodeLlama-70B
CodeLlama-70B
OpenMathInstruct-1: A 1.8 Million Math Instruction Tuning Dataset
NeurIPS 2024 · Table: model comparison
Instruction and fine-tuning
564.7MAS-Architect
GPT-4o-mini
MAS-Architect: Declarative Multi-Agent System Design via Separation of Concerns
ICML 2026 · Table 1
LLM agents
664.19SmartFed
LLaMA2-7B
Don't Reinvent the Wheel, Just Realign the Spokes: Resource-Efficient Federated Fine-Tuning via Rank-Wise Expert Assembly
ICML 2026 · Table 1
Instruction and fine-tuning
763.7ToRA-Code-34B
CodeLLaMA-34B
ToRA: A Tool-Integrated Reasoning Agent for Mathematical Problem Solving
ICLR 2024 · Table: Results on 10 mathemati
Agents, code and math
862.7Q-Function PPO
Codellama Base
Learning Goal-Conditioned Representations for Language Reward Models
NeurIPS 2024 · Table: Policy model performanc
RL from feedback
962.4MAS-GPT
Qwen2.5-Coder-32B-Instruct
MAS-GPT: Training LLMs to Build LLM-based Multi-Agent Systems
ICML 2025 · Table: Comparing MAS-GPT with
LLM agents
1051.43ECONFrom Debate to Equilibrium: Belief‑Driven Multi‑Agent LLM Reasoning via Bayesian Nash Equilibrium
ICML 2025 · Table: Performance of differen
Games and equilibria
5 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.