AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

RiskPO: Risk-based Policy Optimization with Verifiable Reward for LLM Post-Training

Tao Ren, Jinyang Jiang, Hui Yang, Wan Tian, Minhao Zou, Guanghao Li, Zishi Zhang, Qinghao Wang, Shentao Qin, Yanjun Zhao, Rui Tao, Hui Shao and 1 more

ICLR 2026 · Reinforcement learning › RL from feedback · approach: GRPO stability and variants

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
replace mean-based GRPO objective with MVaR risk objectiveGRPO · Hard-level math avg (6 benchmarks)Pass@1 avg: 40.41 -> 46.65Improved
replace mean objective with MVaR vs DAPODAPO · Hard-level math avgPass@1 avg: 43.87 -> 46.65Improved
add bundling of 5 questions (vs no bundling B=1) (ablation)RiskPO · Easy-level math avg (MATH, GSM8K)Pass@1 avg: 65.80 -> 68.25Improved
increase bundle size from 5 to 10 (ablation)RiskPO · Easy-level math avg (MATH, GSM8K)Pass@1 avg: 68.25 -> 66.65Hurt
use bundle size 2 instead of 5 (ablation)RiskPO · Easy-level math avg (MATH, GSM8K)Pass@1 avg: 68.25 -> 67.20Hurt
use risk-seeking objective instead of risk-averse (ablation)RiskPO · MATHPass@1 / entropy: 52% -> 54% vs 52% -> 56%Hurt
lower alpha from 0.2 to 0.1 (ablation)RiskPO · Easy-level math avg (MATH, GSM8K)Pass@1 avg: 68.25 -> 66.90Hurt
raise beta from 0.8 to 0.9 (ablation)RiskPO · Easy-level math avg (MATH, GSM8K)Pass@1 avg: 68.25 -> 66.95Hurt
set omega to 0 (no extra lower-tail weight) (ablation)RiskPO · Easy-level math avg (MATH, GSM8K)Pass@1 avg: 68.25 -> 67.20Hurt
set omega to 1.0 (ablation)RiskPO · Easy-level math avg (MATH, GSM8K)Pass@1 avg: 68.25 -> 67.85Hurt
use RiskPO instead of GRPOGRPO · Pass@k on AMC and MATH500Pass@8/16Improved
use RiskPO instead of GRPOGRPO · DAPOMATH-17K trainingpolicy entropyImproved

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