Unverified50% confidenceFactExact time
The generalization capability of LLMs in temporal extraction tasks cannot be predicted from any single dimension alone, nor can it be reliably inferred from in-domain evaluation or single-dimension evaluation.
1
Sources
50%
Confidence
Long-term
Relevance
10/5/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedLLM output is non-deterministic; due to the non-associativity of floating-point operations and differences in context windows, even with identical inputs, outputs are difficult to reproduce exactly and do not possess idempotency.73% similarUnverified浮点数在token空间中并非连续分布，导致LLM精确运算天然不可靠69% similarUnverified纯LLM探索模式在确定性任务上成功率难以保障，因每次推理存在随机性68% similarUnverifiedTraditional accuracy metrics fail to capture the internal logical contradictions of LLMs, making it necessary to introduce new evaluation standards such as the information processing gap.67% similarUnverifiedMethods such as fuzzing, static analysis, and LLM-as-verifier struggle to cover all possible edge cases.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/975554API
curl https://kongchang.com/api/v1/knowledge/claims/975554MCP
get_claim(id=975554)