Unverified80% confidenceFactTime unknown

上下文控制在100K token以内时模型表现最为稳定，超过100K后幻觉率明显上升

1

Sources

80%

Confidence

Medium-term (~90 days)

Relevance

5/31/2026

First Seen

Valid until: 8/29/2026

Sources

4×3080Ti本地部署千问3.6 27B跑OpenCode编程实测

bilibili苏不二师兄

Related Entities

千问3.6 27Bmodel

Related Claims

Verified思维链效果在参数量超过100B的模型上尤为显著70% similar Unverified当前长上下文窗口（100K-1M tokens）的普及使Agent能够维持复杂任务的长期记忆68% similar UnverifiedH100内存子系统的动态功耗在大模型推理中占比可达整卡功耗的20%～35%68% similar UnverifiedClaude系列模型支持100K+ tokens的长上下文理解能力67% similar Unverified打击类动作对延迟最为敏感，10ms 即可被察觉；而连续控制类动作的容忍度可达 100ms 以上65% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/4401

API

curl https://kongchang.com/api/v1/knowledge/claims/4401

MCP

get_claim(id=4401)