AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

Trust-Region Adaptive Policy Optimization

Mingyu Su, Jian Guan, Yuxian Gu, Minlie Huang, Hongning Wang

ICLR 2026 · Reinforcement learning › RL from feedback · approach: RL for LLM reasoning · 12 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
add micro-group sampling (ablation)GRPO · Math avg (5 benchmarks)accuracy: 50.4 -> 52.7Improved
add TrSFT loss on expert prefixes (ablation)GRPO + micro-group sampling · Math avg (5 benchmarks)accuracy: 52.7 -> 56.6Improved
add standard SFT loss on expert prefixes instead of TrSFT (ablation)GRPO + micro-group sampling · Math avg (5 benchmarks)accuracy: 52.7 -> 32.3Hurt
use LUFFY offline RL loss instead of TrSFT (ablation)GRPO + micro-group sampling · Math avg (5 benchmarks)accuracy: 52.7 -> 53.6Mixed
interleave TRAPO instead of two-stage SFT-then-RLQwen2.5-Math-7B · Math avg (5 benchmarks)accuracy: 54.3 -> 56.6Improved
TRAPO instead of pure GRPOQwen2.5-Math-7B · Math avg (5 benchmarks)accuracy: 50.4 -> 56.6Improved
TRAPO instead of SFTQwen2.5-Math-7B · Math avg (5 benchmarks)accuracy: 50.3 -> 56.6Improved
TRAPO instead of SFT-then-RLQwen2.5-Math-7B · General avg (ARC-c, MMLU-Pro)accuracy: 44.5 -> 68.3Improved
make first micro-group receive 0.1 expert prefix (ablation)TRAPO · Math avg (six benchmarks, 100 steps)accuracy: 39.4 -> 34.1Hurt
add full expert trajectory (L4=1.0) in final micro-group (ablation)TRAPO · Math avg (six benchmarks, 100 steps)accuracy: 37.6 -> 39.4Improved
increasing instead of constant thresholds (ablation)TRAPO · Math avg (six benchmarks, 100 steps)accuracy: 38.1 -> 39.4Improved
increasing instead of decreasing thresholds (ablation)TRAPO · Math avg (six benchmarks, 100 steps)accuracy: 37.4 -> 39.4Improved

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