Unverified50% confidenceTradeoffExact time
If 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.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWhen prompts are too long, large models experience an 'attention explosion' phenomenon, where excessively long inputs cause effective information to be lost.79% similarVerifiedOverly verbose, contradictory, or ambiguous system prompts cause the model's attention to become scattered, resulting in over-thinking or instruction-following failures; streamlining agents.md can improve the determinism of instruction following.78% similarUnverifiedSegments that are too long will dilute relevant information and increase noise, while segments that are too short lack sufficient context, leading to inadequate generation context.78% similarUnverifiedInjecting all chat history into the context leads to problems such as context explosion, attention dilution, and memory fragmentation.78% similarUnverified过度膨胀的上下文会稀释模型注意力、抬高推理成本，甚至引入噪声导致幻觉77% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1011575API
curl https://kongchang.com/api/v1/knowledge/claims/1011575MCP
get_claim(id=1011575)