Unverified50% confidenceFactExact time
In large language model training, a tiny error often occurs well before any anomaly appears in the loss curve.
1
Sources
50%
Confidence
Long-term
Relevance
10/4/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe training objective of large language models is to minimize the prediction loss of the next token (cross-entropy loss), not to verify the factual accuracy of generated content.78% similarUnverified仅凭相对少量的针对性训练数据，就能让大语言模型对明确事实给出完全错误的回答78% similarVerified仅用成功轨迹训练的模型往往存在分布外泛化的严重缺陷，遇到工具返回错误时便丧失纠偏能力78% similarUnverifiedForcing small language models to approximate deterministic problems probabilistically sacrifices accuracy and energy consumption, and much of the unreliability in edge small models is avoidable.76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/972937API
curl https://kongchang.com/api/v1/knowledge/claims/972937MCP
get_claim(id=972937)