SCOPE-RL: Optimizing Reasoning Paths Before and After Success
Reinforcement learning with verifiable rewards (RLVR) optimizes LLMs using sparse verifiable final-answer rewards. This sparse anchor reliably verifies whether a trajectory succeeds but provides no direct feedback on the reasoning path that produced it. Before success, prerequisite progress on hard problems receives no reward signal; after success, outcome rewards cannot distinguish well-organized correct trajectories from redundant or locally flawed ones. We introduce SCOPE-RL (Scaffolded Chain Optimization with Process Efficiency), a two-stage framework that densifies this anchor while retai
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 57%AgileRL/AgileRL →
- PossiblePossibly related (embedding) · 56%hscspring/rl-llm-nlp →
- PossiblePossibly related (embedding) · 54%rllm-org/rllm →
- PossiblePossibly related (embedding) · 50%RL without TD learning →
- PossiblePossibly related (embedding) · 49%RLHF →
- LinkedLinked via arxiv author · 85%Xiaojian Liu →
“SCOPE-RL: Optimizing Reasoning Paths Before and After Success”
- LinkedLinked via arxiv author · 85%Zihan Xu →
“SCOPE-RL: Optimizing Reasoning Paths Before and After Success”
- LinkedLinked via arxiv author · 85%Jianqiang Xia →
“SCOPE-RL: Optimizing Reasoning Paths Before and After Success”