Unverified50% confidenceFactExact time
A major pain point of large language models in mathematics is that they may generate proofs that appear rigorous but actually contain skipped steps or errors
1
Sources
50%
Confidence
Long-term
Relevance
10/6/2026
First Seen
Sources
Related Entities
Related Claims
Unverified大语言模型在数学证明、代码生成等任务上超越人类平均水平，却在空间推理、计数等简单任务上出现失误，呈现非均匀能力分布78% similarVerifiedLarge language models carry a hallucination risk in numerical reasoning and may produce errors in scenarios such as aggregate calculations, time-range filtering, or multi-table joins.76% similarUnverified引导过程若缺乏严格的数学保证，可能引入难以察觉的系统性误差75% 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.74% similarUnverifiedIn large language model training, a tiny error often occurs well before any anomaly appears in the loss curve.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/982567API
curl https://kongchang.com/api/v1/knowledge/claims/982567MCP
get_claim(id=982567)