Verified65% confidenceFactExact time
Current mainstream large language models exhibit significant deficiencies in both confidence calibration and selective abstention, tend to be overconfident, and when faced with questions outside the training distribution, tend to produce fluent but incorrect answers.
3
Sources
65%
Confidence
Medium-term (~90 days)
Relevance
9/28/2026
First Seen
Valid until: 12/27/2026
Sources
Related Entities
Related Claims
UnverifiedThe trade-off between training efficiency and inference quality in diffusion language models is not yet ideal; many of the results shown depend on carefully designed experimental conditions, and a gap remains in performance on open-domain tasks.80% similarUnverifiedThe confidence level of a language model and the reliability of its judgments are often not correlated, posing the risk of hallucinatory overconfidence.79% similarUnverifiedThe persuasiveness trap of large language models is essentially a calibration failure, as there is no positive correlation between output fluency and factual accuracy.78% similarVerifiedWhen a model lacks accurate knowledge of a particular library, it tends to generate syntactically correct but semantically incorrect code—so-called 'confident errors.'76% similarUnverifiedThe handling of negation by embedding models is a systematic flaw, because mainstream training objectives (masked language modeling, contrastive learning) do not explicitly require the model to distinguish between positive and negative semantics, causing 'like' and 'dislike' to be encoded at nearby positions.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/955413API
curl https://kongchang.com/api/v1/knowledge/claims/955413MCP
get_claim(id=955413)