1 paper · 1 filter
Pawan K. Tripathi, Hemant Sharma, Andrew Chuang +1
LLM agents can make expert scientific workflows accessible through natural language, but a plausible response does not establish that the requested computation ran. This matters at…