ConSPO introduces a contrastive sequence-level policy optimization that aligns rollout scores with generation likelihoods via length-normalized log-probabilities and an InfoNCE-style group contrast with curriculum margin to outperform GRPO on LLM math reasoning benchmarks.
arXiv preprint arXiv:2512.13095 , year=
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 3verdicts
UNVERDICTED 3roles
background 1polarities
background 1representative citing papers
PruneTIR prunes erroneous tool-call trajectories during LLM inference via three trigger-based components to raise Pass@1 accuracy and efficiency while shortening context.
Prefix Sampling replays self-generated trajectory prefixes to control rollout pass rates near 50% in binary-reward RL, delivering wall-clock speedups and modest performance gains on SWE-bench Verified and AIME tasks.
citing papers explorer
-
Revisiting Reinforcement Learning with Verifiable Rewards from a Contrastive Perspective
ConSPO introduces a contrastive sequence-level policy optimization that aligns rollout scores with generation likelihoods via length-normalized log-probabilities and an InfoNCE-style group contrast with curriculum margin to outperform GRPO on LLM math reasoning benchmarks.
-
PruneTIR: Inference-Time Tool Call Pruning for Effective yet Efficient Tool-Integrated Reasoning
PruneTIR prunes erroneous tool-call trajectories during LLM inference via three trigger-based components to raise Pass@1 accuracy and efficiency while shortening context.
-
Rollout Pass-Rate Control: Steering Binary-Reward RL Toward Its Most Informative Regime
Prefix Sampling replays self-generated trajectory prefixes to control rollout pass rates near 50% in binary-reward RL, delivering wall-clock speedups and modest performance gains on SWE-bench Verified and AIME tasks.