Tag
LAPO proposes a leave-one-turn attribution method for self-generated process rewards in multi-turn search reasoning, enabling fine-grained credit assignment without external reward models. It achieves state-of-the-art results across seven datasets.
IdeaTrail is a dataset of multi-turn process trajectories for scientific ideation, synthesizing research processes from evidence gathering to proposal construction using a Generator–Advisor loop to ensure grounding.
SkillCoach introduces a self-evolving rubric framework that evaluates and enhances LLM agent skill-use by analyzing skill selection, following, composition, and reflection, providing process-level supervision beyond outcome-only metrics.
PASS is a middleware that fixes three pathologies in process-supervised RL for LLM reasoners, improving GRPO by independently standardizing streams, chunking by value, and using average value density. It shows consistent gains in math reasoning and multi-hop QA.
VeriGate extends GRPO with verifier-gated step-level supervision, providing fine-grained credit assignment when verifier rewards are degenerate. It achieves substantial accuracy improvements on reasoning benchmarks for 1.5B and 7B models.
CorVer is a lightweight, corpus-grounded reward mechanism that uses Wikipedia co-occurrence statistics to provide efficient sentence-level feedback for reinforcement learning in factual question answering, outperforming neural verifiers while training 4.8 to 8.4x faster.
STRIDE introduces a training framework that uses learnable stepwise language feedback instead of scalar rewards to improve LLM reasoning, achieving state-of-the-art results on diverse benchmarks.
Introduces IOP, a framework that internalizes outcome supervision into process supervision for reasoning reinforcement learning, enabling fine-grained credit assignment without external annotations.
ATTNPO introduces an attention-guided process supervision framework that reduces overthinking in large reasoning models by leveraging intrinsic attention signals for step-level credit assignment, achieving improved performance with shorter reasoning lengths across 9 benchmarks.
OpenAI demonstrates that process supervision—rewarding intermediate reasoning steps rather than just final answers—improves mathematical reasoning while reducing alignment costs. This approach produces more interpretable, human-aligned reasoning without sacrificing model performance.