Search papers, labs, and topics across Lattice.
The paper introduces GBQA, a new benchmark for evaluating LLMs' ability to autonomously detect software bugs in games, a challenging domain due to its dynamic runtime environments. GBQA contains 30 games with 124 human-verified bugs across three difficulty levels, generated using a multi-agent system with human oversight. Experiments using a ReAct-based interactive agent on frontier LLMs reveal that even the best model, Claude-4.6-Opus, only identifies 48.39% of the bugs, highlighting the difficulty of autonomous bug discovery.
Even state-of-the-art LLMs struggle to find half the bugs in a new game QA benchmark, revealing a significant gap in autonomous software engineering capabilities.
The autonomous discovery of bugs remains a significant challenge in modern software development. Compared to code generation, the complexity of dynamic runtime environments makes bug discovery considerably harder for large language models (LLMs). In this paper, we take game development as a representative domain and introduce the Game Benchmark for Quality Assurance (GBQA), a benchmark containing 30 games and 124 human-verified bugs across three difficulty levels, to evaluate whether LLMs can autonomously detect software bugs. The benchmark is constructed using a multi-agent system that develops games and injects bugs in a scalable manner, with human experts in the loop to ensure correctness. Moreover, we provide a baseline interactive agent equipped with a multi-round ReAct loop and a memory mechanism, enabling long-horizon exploration of game environments for bug detection across different LLMs. Extensive experiments on frontier LLMs demonstrate that autonomous bug discovery remains highly challenging: the best-performing model, Claude-4.6-Opus in thinking mode, identifies only 48.39% of the verified bugs. We believe GBQA provides an adequate testbed and evaluation criterion, and that further progress on it will help close the gap in autonomous software engineering.
