DecompRL: Solving Harder Problems by Learning Modular Code Generation
How can Large Language Models (LLMs) solve problems they currently cannot? Repeated sampling scales test-time compute but GPU cost grows linearly with attempts, while reinforcement learning (RL) with verifiable rewards improves single-attempt accuracy at the expense of sample diversity. Both strategies ultimately fail when the base policy has near-zero probability of producing a correct solution: no amount of sampling or gradient signal can overcome a search space that is simply too large. We take a different approach: rather than sampling harder, we make the task easier by decomposing problem
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) · 54%rllm-org/rllm →
- PossiblePossibly related (embedding) · 53%chrisliu298/awesome-llm-unlearning →
- PossiblePossibly related (embedding) · 52%IEEE Rolls Out Large Language Models Virtual Training Course →
- LinkedLinked via arxiv author · 85%Juliette Decugis →
“DecompRL: Solving Harder Problems by Learning Modular Code Generation”
- LinkedLinked via arxiv author · 85%Fabian Gloeckle →
“DecompRL: Solving Harder Problems by Learning Modular Code Generation”
- LinkedLinked via arxiv author · 85%Francis Bach →
“DecompRL: Solving Harder Problems by Learning Modular Code Generation”
- LinkedLinked via arxiv author · 85%Taco Cohen →
“DecompRL: Solving Harder Problems by Learning Modular Code Generation”
- LinkedLinked via arxiv author · 85%Gabriel Synnaeve →
“DecompRL: Solving Harder Problems by Learning Modular Code Generation”