Expired75% confidenceFactTime unknown
An evaluation report published by Kwan Ma's team in July 2025 shows that as the number of input tokens increases, nearly all mainstream models experience performance degradation on simple tasks.
1
Sources
75%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026(expired)
Sources
AI Agent模型分层路由策略：Token成本、幻觉应对与选型实战
bilibili前端小卒
Related Entities
Related Claims
Partially VerifiedA 2023 study by Stanford University's Nelson Liu and colleagues found that when key information is placed in the middle of a long context, model performance can drop by more than 20%.72% similarUnverified大模型的性能提升已经明显放缓，最近新版本的评分甚至不一定超越之前的模型，能力基本已定型65% similarUnverified研究表明，即使在 GSM8K 或 MATH 数据集上接近满分的模型，面对稍作变形或需要跨领域组合的问题时性能下降可达30-50个百分点64% similarUnverifiedStanford University's 2023 "Lost in the Middle" study (Liu et al.) showed that when key information is located in the middle section of the context, the model's retrieval accuracy drops approximately 20–30 percentage points compared to the head and tail, and this is especially significant when the context exceeds 32K tokens.64% similarUnverified研究表明在超过8-10轮对话后，模型对初始系统提示中规则的遵循率可下降20%-40%63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/27302API
curl https://kongchang.com/api/v1/knowledge/claims/27302MCP
get_claim(id=27302)