Unverified50% confidenceFactExact time

从信息论和香农定理角度看，任何超过信息熵下界的压缩都必然丢失信息，因此无论模型多大，压缩必然是有损的

1

Sources

50%

Confidence

Long-term

Relevance

7/15/2026

First Seen

Sources

Retrieval vs. Reasoning: Does Search Technology Still Matter in the Age of LLMs? Six Experts Debate

Related Claims

Unverified信息论先驱香农指出任何信道容量都存在上限，当信源信息熵超过信道容量时信息损失不可避免78% similar Unverified推理能力可通过知识蒸馏和强化学习微调压缩，但事实性知识的存储依赖参数规模，压缩会导致知识覆盖面下降和幻觉率上升78% similar Unverified香农提出的信息熵概念表明低概率事件携带更高的信息量74% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/517741

API

curl https://kongchang.com/api/v1/knowledge/claims/517741

MCP

get_claim(id=517741)