AI Paper Atlas

Research map / Reinforcement learning / RL from feedback

PoLi-RL: A Point-to-List Reinforcement Learning Framework for Conditional Semantic Textual Similarity

Zixin Song, Bowen Zhang, Qian-Wen Zhang, di yin, Xing Sun, Chunping Li

ICLR 2026 · Reinforcement learning › RL from feedback · approach: RL for LLM reasoning · 2 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 the two-stage pointwise-to-hybrid curriculum instead of naive listwise RL (ablation)Qwen3-8B RL · C-STSSpearman: 38.19 -> 48.18Improved
naive listwise RL instead of few-shot (ablation)Qwen3-8B · C-STSSpearman: 37.9 -> 38.19No clear effect
add binary reward in Stage I (ablation)PoLi-RL Stage I · C-STSSpearman: 44.19 -> 44.77Improved
add Stage II hybrid reward after Stage I (ablation)PoLi-RL · C-STSSpearman: 44.77 -> 48.18Improved
add listwise reward to Stage II (ablation)PoLi-RL Stage II · C-STSSpearman: 46.71 -> 48.18Improved
add pairwise reward to Stage II (ablation)PoLi-RL Stage II · C-STSSpearman: 47.6 -> 48.18Improved
slice size N=24 instead of 16, 32, 40 or 48 (ablation)PoLi-RL PSRR · C-STSSpearman: 47.16 / 47.44 / 47.18 / 46.78 -> 48.18Improved
pairwise weight 1.5 instead of 1.0 (ablation)PoLi-RL Stage II · C-STSSpearman: 47.83 -> 48.18Improved
generation multiplicity G=8 instead of 4 or 12 (ablation)PoLi-RL Stage II · C-STSSpearman: 47.47 / 47.98 -> 48.18Mixed
pairwise base reward 0.5 instead of 0 (ablation)PoLi-RL Stage II · C-STSSpearman: 47.90 -> 48.18Improved
vary binary reward weight lambda2 in Stage I (ablation)PoLi-RL Stage I · C-STSSpearman: 44.94 / 44.77 / 44.67 / 44.76No clear effect
RL instead of regression SFT with Pcc lossQwen3-8B · C-STSSpearman: 46.44 -> 48.18Improved

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