Scaling Agentic Verifier for Competitive Coding
This addresses the challenge of improving code generation accuracy for competitive programming, though it is incremental as it builds on existing execution-based re-ranking methods.
The paper tackled the problem of LLMs struggling to solve competitive programming problems correctly in a single attempt by proposing Agentic Verifier, an execution-based agent that actively searches for discriminative test inputs, resulting in up to +10-15% absolute gains in Best@K accuracy across benchmarks.
Large language models (LLMs) have demonstrated strong coding capabilities but still struggle to solve competitive programming problems correctly in a single attempt. Execution-based re-ranking offers a promising test-time scaling strategy, yet existing methods are constrained by either difficult test case generation or inefficient random input sampling. To address this limitation, we propose Agentic Verifier, an execution-based agent that actively reasons about program behaviors and searches for highly discriminative test inputs that expose behavioral discrepancies among candidate solutions. Through multi-turn interaction with code execution environments, the verifier iteratively refines the candidate input generator and produces targeted counterexamples rather than blindly sampling inputs. We train the verifier to acquire this discriminative input generation capability via a scalable pipeline combining large-scale data synthesis, rejection fine-tuning, and agentic reinforcement learning. Extensive experiments across five competitive programming benchmarks demonstrate consistent improvements over strong execution-based baselines, achieving up to +10-15% absolute gains in Best@K accuracy. Further analysis reveals clear test-time scaling behavior and highlights the verifier's broader potential beyond reranking.