Unverified50% confidenceFactExact time
Research shows that over-refusal not only lowers task completion rates but also reduces users' trust in the system, and after repeatedly hitting walls they give up using it or switch to competing products.
1
Sources
50%
Confidence
Long-term
Relevance
11/7/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedExcessive safety fine-tuning can cause models to over-refuse, undermining their practical utility.76% similarUnverifiedThe benefit of end-to-end generation is a low barrier to use and reproducible processes, but it also means troubleshooting is harder when something goes wrong, and users have less control over the intermediate process.73% similarVerifiedOver-refusal strategies reduce risk but harm model utility, especially in scenarios such as professional development and security research that require touching sensitive boundaries.71% similarUnverified天花板效应指当测试工具的难度上限低于被测对象的实际能力时,分数分布呈现严重负偏态,导致测试的信度和效度显著下降71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1147081API
curl https://kongchang.com/api/v1/knowledge/claims/1147081MCP
get_claim(id=1147081)