HFEPX · Eval paper review
Haipeng Luo, Qingfeng Sun, Songli Wu, Can Xu +3 more
Published
Jun 17, 2026
Citations
0
Trust level
Provisional
Usefulness score
Unavailable
Extraction confidence
0% (Provisional)
Derived from abstract and metadata only.
Signals refreshed
Jun 17, 2026
Signal extraction is still processing. This page currently shows metadata-first guidance until structured protocol fields are ready.
This page is a lightweight research summary built from the abstract and metadata while deeper extraction catches up.
All signals on this page are inferred from the abstract only and may be inaccurate. Do not use this page as a primary protocol reference.
Best use
Background context only
Use if you need
A provisional background reference while structured extraction finishes.
What to verify
Read the full paper before copying any benchmark, metric, or protocol choices.
Main weakness
This page is still relying on abstract and metadata signals, not a fuller protocol read.
Eval-fit score is unavailable until extraction completes.
If you are doing eval pipeline work, start here
Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training. We conduct a first-order gradient analysis of token-level entropy dynamics under GRPO and identify a token-level credit assignment mismatch: the per-token entropy variation decomposes into the product of the trajectory-level advantage and an entropy sensitivity function over the next-token distribution, yielding an advantage-surprisal four-quadrant structure and a near-criticality property. Motivated by it, we propose STARE (Surprisal-guided Token-level Advantage Reweighting for policy Entropy stability), which identifies entropy-critical token subsets via batch-internal surprisal quantiles, selectively reweights their effective advantages, and incorporates a target-entropy closed-loop gate for stable entropy regulation. Across model scales from 1.5B to 32B and three task families (Short CoT, Long CoT, and Multi-Turn Tool Use), STARE sustains stable RL training over thousands of steps while maintaining policy entropy within the target band. On AIME24 and AIME25, STARE outperforms DAPO and other competitive baselines by 4%-8% in average accuracy, with reflection tokens and response length growing in tandem, indicating sustained exploration-exploitation balance that further unlocks RL training potential.Code is available at https://github.com/hp-luo/STARE.
These are the protocol signals we could actually recover from the available paper metadata. Use them to decide whether this paper is worth deeper reading.
None explicit
No explicit feedback protocol extracted.
"Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training."
Automatic metrics, Tool Use evaluation
Includes extracted eval setup.
"Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training."
Not reported
No explicit QC controls found.
"Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training."
Not extracted
No benchmark anchors detected.
"Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training."
Accuracy
Useful for evaluation criteria comparison.
"On AIME24 and AIME25, STARE outperforms DAPO and other competitive baselines by 4%-8% in average accuracy, with reflection tokens and response length growing in tandem, indicating sustained exploration-exploitation balance that further unlocks RL training potential.Code is available at https://github.com/hp-luo/STARE."
Unknown
Rater source not explicitly reported.
"Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training."
This page is using abstract-level cues only right now. Treat the signals below as provisional.
Evaluation fields are inferred from the abstract only.
Reinforcement Learning with Verifiable Rewards algorithms like GRPO have emerged as the dominant post-training paradigm for complex reasoning in LLMs, yet commonly suffer from policy entropy collapse during training.
Based on abstract + metadata only. Check the source paper before making high-confidence protocol decisions.
