SCRL derives verifiable subproblems from reference reasoning chains and applies subproblem-level normalization to enable finer-grained credit assignment in RLVR for LLM math reasoning.
When the ball bounces, it travels in a straight line in this grid of triangles
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
From Reasoning Chains to Verifiable Subproblems: Curriculum Reinforcement Learning Enables Credit Assignment for LLM Reasoning
SCRL derives verifiable subproblems from reference reasoning chains and applies subproblem-level normalization to enable finer-grained credit assignment in RLVR for LLM math reasoning.