Verified75% confidenceFactExact time
Context window size does not equal effective context length; exceeding the effective attention range causes a forgetting drift phenomenon.
3
Sources
75%
Confidence
Long-term
Relevance
7/10/2026
First Seen
Sources
GLM-5.2发布：开源编程能力首次逼近闭源旗舰模型
bilibili五里墩茶社6/22/2026
Related Claims
UnverifiedSimply expanding the context window does not equate to a linear improvement in comprehension ability; there is a phenomenon of forgetting in the middle.80% similarUnverified更长的上下文窗口并未消除注意力稀释效应，反而带来更高的单次调用成本79% similarUnverifiedIf fact extraction granularity is too coarse, it leads to overgeneralization and false triggering; if too fine, it cannot be reused across tasks, degrading the memory system into a simple dialogue cache.76% similarUnverifiedWhen the context window is nearly full, context compression is performed by discarding irrelevant information or generating summaries, in order to control token costs and keep the model's attention focused.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/438700API
curl https://kongchang.com/api/v1/knowledge/claims/438700MCP
get_claim(id=438700)