AI Paper Atlas

Research map / Reinforcement learning / Imitation learning

GAR: Generative Adversarial Reinforcement Learning for Formal Theorem Proving

Ruida WANG, Jiarui Yao, Rui Pan, Shizhe Diao, Tong Zhang

ICLR 2026 · Reinforcement learning › Imitation learning · approach: Imitation combined with RL · 6 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
apply GAR trainingGoedel-Prover-V2-8B · MiniF2F-Test pass@32pass rate: 77.87% -> 80.33%Improved
apply GAR trainingDeepSeek-Prover-V2-7B · MiniF2F-Test pass@32pass rate: 70.49% -> 74.18%Improved
apply GAR trainingDeepSeek-Prover-V2-7B · ProofNet-Testpass rate: 22.58% -> 25.81%Improved
apply GAR trainingDeepSeek-Prover-V2-7B · PutnamBench pass@32problems solved: 22/660 -> 24/660Improved
add the statement modification penalty (ablation)GAR · Statement modification rate, step 4modification rate: 74.11% -> 33.63%Improved
use GAR instead of further GRPO on static data (ablation)Goedel-Prover-V2-8B · MiniF2F-Test pass@32pass rate: 77.46% -> 80.33%Improved
add further GRPO on static data (ablation)Goedel-Prover-V2-8B · MiniF2F-Test pass@32pass rate: 77.87% -> 77.46%No clear effect
train the statement fuser too (not frozen) (ablation)GAR on Goedel-Prover-V2-8B · MiniF2F-Test pass@32pass rate: 77.87% -> 80.33%Improved
use problem fusion instead of single-problem MagicCoder evolving (ablation)GAR on Goedel-Prover-V2-8B · MiniF2F-Test pass@32pass rate: 75.41% -> 80.33%Improved
use GAR-trained prover instead of base prover to answer fused statementsstatement fuser study · Proof correctness iteration 4average proof correctness rate: 7.69% -> 21.79%Improved
apply GAR trainingDeepSeek-Prover-V2 · MiniF2F-Test modification ratestatement modification rate: 6.96% -> 13.11%Hurt
apply GAR trainingGoedel-Prover-V2 · MiniF2F-Test modification ratestatement modification rate: 24.18% -> 27.05%Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy