Unverified50% confidenceOpinionExact time
The paper's analysis points out that when judging "when to ask," models tend to act too early or too late, and on "what to ask," they often pose irrelevant or redundant questions.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/30/2026
First Seen
Valid until: 1/28/2027
Sources
Related Entities
Related Claims
UnverifiedClarifying follow-up questions are considered the most direct strategy for handling intent uncertainty, meaning the model proactively asks for clarification when its confidence falls below a threshold rather than rashly calling a tool.70% similarUnverifiedHow to make the model accurately judge which requests are worth reasoning about is itself a technical challenge70% similarVerifiedWhen a model gives answers of vastly different quality for semantically equivalent but differently phrased questions, it indicates that the model has overfitted to surface-level wording rather than truly mastering the underlying task.69% similarUnverified当目标不明确时建议使用Ask问答模式梳理需求，目标明确后切换到执行模式操作69% similarUnverifiedIn complex context scenarios, understanding the question is more important than answering it.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1114227API
curl https://kongchang.com/api/v1/knowledge/claims/1114227MCP
get_claim(id=1114227)