ARIADNE: Agentic Reward-Informed Adaptive Decision Exploration via Blackboard-Driven MCTS for Competitive Program Generation

TL;DR

This paper introduces ARIADNE, a blackboard-driven MCTS framework that improves competitive program generation by modeling it as a sequential decision process with structured evidence accumulation.

Contribution

ARIADNE's novel blackboard-driven MCTS approach enables systematic exploration and effective feedback use, outperforming existing methods in competitive programming benchmarks.

Findings

Achieves up to 26.06 points higher Pass@1 scores than baseline methods.

Consistently outperforms strong baselines across four benchmarks.

Effectively utilizes execution feedback within limited computational budgets.

Abstract

Competitive program generation aims to automatically produce correct and efficient solutions for programming-contest problems under strict time and memory constraints. Existing LLM-based approaches often fail to perform explicit algorithmic planning and to handle edge cases robustly, leading to unreliable one-shot generation. Moreover, although execution feedback is essential for iterative debugging and refinement, incorporating such feedback effectively within limited computational budgets remains difficult. To overcome these limitations, we propose {\tool}, a blackboard-driven Monte Carlo Tree Search (MCTS) framework that models program generation as a sequential decision process. {\tool} organizes the generation workflow into five coordinated stages (i.e., strategy selection, code generation, test generation, quality evaluation, and code repair) while maintaining a shared blackboard that accumulates structured evidence to guide subsequent decisions. Experiments on four benchmarks (APPS, CodeContests, CodeContests+, and LiveCodeBench) show that {\tool} consistently achieves the best Pass@1 performance across multiple LLM backends. With GPT-4o, {\tool} attains Pass@1 scores of 41.30, 46.67, 27.27, and 20.91, surpassing the strongest baseline CodeSim by up to 26.06 points, while further improvements are observed with DeepSeek-V3.2. These results indicate that combining global search through MCTS with persistent evidence accumulation on a shared blackboard enables systematic exploration and effective feedback utilization, substantially enhancing the capability of LLMs in competitive program generation.