2026
HARDTESTGEN: A High-Quality RL Verifier Generation Pipeline for LLM Algorithimic Coding
ICLR 2026poster
Verifiers provide important reward signals for reinforcement learning of large language models (LLMs). However, it is challenging to develop or create reliable verifiers, especially for code generation tasks. A well-disguised wrong solution program may only be detected by carefully human-written edg…