Unverified50% confidenceFactExact time
The format line directly determines how often the model shows its reasoning process: "Answer with the number only" led to 37% of responses showing reasoning, the plain version 27%, and the florid version only 8%.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe oracle router knows in advance which cases the small model will get wrong, sends only the wrong requests to the large model, and achieves accuracy comparable to always using the large model at only 29% of the cost.76% similarUnverifiedRanked by question difficulty, persistence rate was 91.4% on the 60 questions that all models answered correctly, but only 73.5% on the 52 questions that at least one model got wrong, indicating that sycophancy tracks uncertainty.74% similarUnverifiedThe impact of reasoning-chain corruption on final answers was relatively small (2%–30%), and only 2.9%–35.3% of support-judgment flips co-occurred with answer changes.72% similarUnverifiedIn a test of 12 models across 112 questions, when faced with a "Are you sure?" challenge, models had a 94.4% probability of maintaining their original correct answer.72% similarUnverified根据文中量化估算，编程智能体任务中模型推理仅占总时间的 20%-30%，剩余 70%-80% 花在等待外部系统 I/O 响应71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1018309API
curl https://kongchang.com/api/v1/knowledge/claims/1018309MCP
get_claim(id=1018309)