Zhi Jin

PRAXIS reveals that systematically extracting and surfacing tacit knowledge can dramatically enhance LLM performance in domain-specific code generation, outperforming traditional methods.

General LLMs struggle with enzyme classification, but leveraging external knowledge can dramatically improve their performance, revealing hidden gaps in reasoning capabilities.

Relevance estimation in e-commerce can be dramatically improved by treating it as a sequential decision process, leading to a 15.94% reduction in bad-case rates.

Strong code generation doesn't guarantee effective requirement clarification, exposing a critical flaw in LLM capabilities that could hinder software engineering practices.

Intent-driven rule completion in IoT systems can boost completion rates by 43% while embedding safety and traceability throughout the process.

MAAD not only automates architecture design but also enhances the quality of outputs through a collaborative agent framework and advanced LLM integration.

Ditch the one-size-fits-all code intelligence: modeling individual developer behaviors inside the IDE boosts Q&A accuracy by 33.8%.

LLMs can significantly improve code generation by interleaving reasoning steps *within* the code itself, adaptively focusing on high-complexity areas.