Unverified50% confidenceFactExact time
The research team evaluated three DRAs and found that when user queries are underspecified, the degree to which agent answers satisfy users' information needs drops significantly.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedA new study systematically evaluated the impact of query ambiguity on agents' final answer quality and intermediate search behavior through the UNDERSPEC-DR framework.75% similarUnverifiedWhen facing ambiguous queries, DRAs conduct broader searches, but these searches rely less on the original user query and more on prior knowledge the model learned during training.74% similarUnverifiedExperiments show that non-agent methods frequently miss changes requested by users while also disturbing regions that should remain unchanged.69% similarUnverifiedThe evaluation challenge for an Agent that reads purchase signals from web text is essentially a machine learning evaluation problem combining weak supervision, delayed feedback, and sample bias, rather than a purely prompt engineering problem.67% similarUnverifiedFor long-tail queries — highly personalized questions that are rarely repeated — the benefit of knowledge reuse systems approaches zero or even becomes negative.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1113924API
curl https://kongchang.com/api/v1/knowledge/claims/1113924MCP
get_claim(id=1113924)