[KongchangAI]
Unverified50% confidenceBenchmarkExact time

A linear probe classifying the selection arm achieves an accuracy of up to 59.7% (chance level 50%), indicating that the selection behavior leaves detectable traces in the model's representations.

1
Sources
50%
Confidence
Long-term
Relevance
10/21/2026
First Seen

Sources

Related Entities

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/1079356
API
curl https://kongchang.com/api/v1/knowledge/claims/1079356
MCP
get_claim(id=1079356)