Ensuring the correctness of LLM-generated code is a core challenge for modern software engineering. Benchmarks for agentic code generation check correctness with held-out test suites, which are inherently incomplete and increasingly susceptible to memorization. Formal verification avoids both problems, but existing wor...
George Ma, Benjamin Mikek, Hao-Yu Li et al.· 0 citations
Fidelity probes are introduced to check a specification against the code and guide its revision, using an observability rule to focus probe questions on behaviour that the modernized system is intended to preserve, including user-visible outputs, changes to stored business data, and interactions with other systems.
Ferhat Erata, Hao Zhou, Jun Huan· arXiv.org· 1 citation
Ensuring the correctness of LLM-generated code is a core challenge for modern software engineering. Benchmarks for agentic code generation check correctness with held-out test suites, which are inherently incomplete and increasingly susceptible to memorization. Formal verification avoids both problems, but existing wor...
George Ma, Benjamin Mikek, Hao-Yu Li et al.· 0 citations
The second KDD Day on AI Reasoning brings together researchers and practitioners from academia and industry to examine how large language models and foundation models can be made more capable, reliable, interpretable, and efficient.
Jun Huan, James Caverlee, Lei Li et al.· Proceedings of the 32nd ACM...· 0 citations
A critical limitation of existing confidence-based VF-TTS methods is demonstrated by showing that such methods catastrophically break down on complex tasks, and a novel selection framework, consilience, is introduced, a novel selection framework that explicitly evaluates the temporal asymmetry of confidence in reasonin...
Lecheng Kong, Like Hui, Hai-Tao Mao et al.· 0 citations
Across four RoboTwin tasks spanning different horizons and coordination patterns, Prism-GRPO improves success and quality at matched rollout budgets and reaches target success rates with up to 56% fewer rollouts.
Zeyun Deng, Yuzhe Lu, Ya-Wei Wang et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
