Checking whether research code matches a paper requires looking in both directions: locating implementations of written claims and finding consequential code behavior the paper may omit. Dude assigns these searches to specialized language-model agents. It also addresses two sources of false alarms: treating one imagined implementation as mandatory, and treating incidental code details as research claims that need documentation. Agents request more groundingTying a model's answers to checkable sources — retrieved documents, live data, tool results — instead of letting it answer from memory alone.Full definition → descriptions and filter findings before producing a report. The authors' experiments show that simply adding agents increased detection while also increasing false alarms; Dude's negotiation and filtering improved that balance. Methods and results.
Terms in this piece · Glossary
grounding — Tying a model's answers to checkable sources — retrieved documents, live data, tool results — instead of letting it answer from memory alone.
multi-agent — Using several AI agents on one problem — splitting work in parallel, checking each other, or filling different roles like planner and reviewer.
AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
articleLabAgent: Customize Any Research Hubs for Scientific Discoveries Using AI AgentsLei Liu, Yikun Zhang, Jialin Chen, Wanjia Zhao, Rex Ying, Wengong Jin, Hua Xu, James Zou, Tianyu Liu, Hongyu Zhao
articleNot All Agents Are Equal: Code Quality and Post-Merge Maintenance Across Five Autonomous Coding Agents in the WildObada Kraishan
articleA Large-Scale Empirical Study of Quality Assurance Practices and Gaps in AI AgentsWuyang Dai, Moses Openja, Jiho Shin, Hung Viet Pham, Song Wang