Language models produce overcomplete reasoning traces where on average 46% of steps can be removed while preserving the answer in 86% of cases, with necessity concentrated in the top three steps.
arXiv preprint arXiv:2506.08745 , year=
10 Pith papers cite this work. Polarity classification is still indexing.
10
Pith papers citing it
representative citing papers
RiVER applies calibrated ranking rewards from execution scores to train LLMs on score-based tasks without ground-truth, producing gains on both heuristic contests and exact-solution coding benchmarks.
CRPO extends group relative policy optimization with stage-dependent uncertainty modeling and reports a 10.4 percentage point weighted F1 gain over RL baselines across 8 mental health datasets.
A parameter-free sampling strategy called CUTS combined with Mixed-CUTS training prevents mode collapse in RL for saturated LLM reasoning tasks and raises AIME25 Pass@1 accuracy by up to 15.1% over standard GRPO.
GeoMin uses geometric distribution modeling on labeled data to assess self-reward reliability, enabling better performance in semi-supervised RLVR with only 10% of typical annotations.
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
RLAVR uses the Corrective Advantage Gap metric and CARE policy to actively acquire ground-truth labels for key samples, stabilizing RLVR training and boosting performance with limited annotation budgets.
ConsistRM improves generative reward models via consistency-aware self-training, outperforming vanilla RFT by 1.5% on average across five benchmarks and four base models.
PivotTrace selects unlabeled data for RLVR by quantifying uncertainty via pivot density from attention dynamics, outperforming full supervision using only 29.3% annotations and converging 2.75 times faster.
The paper delivers the first systematic review of self-evolving agents, structured around what components evolve, when adaptation occurs, and how it is implemented.
citing papers explorer
-
Uncovering the Representation Geometry of Minimal Cores in Overcomplete Reasoning Traces
Language models produce overcomplete reasoning traces where on average 46% of steps can be removed while preserving the answer in 86% of cases, with necessity concentrated in the top three steps.
-
Reinforcement Learning without Ground-Truth Solutions can Improve LLMs
RiVER applies calibrated ranking rewards from execution scores to train LLMs on score-based tasks without ground-truth, producing gains on both heuristic contests and exact-solution coding benchmarks.
-
Mental-R1: Aligning LLM Reasoning for Mental Health Assessment
CRPO extends group relative policy optimization with stage-dependent uncertainty modeling and reports a 10.4 percentage point weighted F1 gain over RL baselines across 8 mental health datasets.
-
Too Correct to Learn: Reinforcement Learning on Saturated Reasoning Data
A parameter-free sampling strategy called CUTS combined with Mixed-CUTS training prevents mode collapse in RL for saturated LLM reasoning tasks and raises AIME25 Pass@1 accuracy by up to 15.1% over standard GRPO.
-
GeoMin: Data-Efficient Semi-Supervised RLVR via Geometric Distribution Modeling
GeoMin uses geometric distribution modeling on labeled data to assess self-reward reliability, enabling better performance in semi-supervised RLVR with only 10% of typical annotations.
-
Trust Region On-Policy Distillation
TrOPD stabilizes on-policy distillation for LLMs with trust-region learning, outlier estimation, and off-policy guidance, outperforming prior OPD methods on reasoning and code benchmarks.
-
When Self-Belief Misleads: Active Label Acquisition for Reinforcement Learning with Verifiable Rewards
RLAVR uses the Corrective Advantage Gap metric and CARE policy to actively acquire ground-truth labels for key samples, stabilizing RLVR training and boosting performance with limited annotation budgets.
-
ConsistRM: Improving Generative Reward Models via Consistency-Aware Self-Training
ConsistRM improves generative reward models via consistency-aware self-training, outperforming vanilla RFT by 1.5% on average across five benchmarks and four base models.
-
Smart Picks in the Dark: Towards Efficient RLVR for Reasoning via Tracing Metacognitive Pivots
PivotTrace selects unlabeled data for RLVR by quantifying uncertainty via pivot density from attention dynamics, outperforming full supervision using only 29.3% annotations and converging 2.75 times faster.
-
A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence
The paper delivers the first systematic review of self-evolving agents, structured around what components evolve, when adaptation occurs, and how it is implemented.