[KongchangAI]
Unverified50% confidenceFactExact time

The version-coupling problem of LLM-as-judge: the judge model itself is also updated with versions, and its scoring criteria and tolerance for wording may shift; when the evaluated model and the judge model are upgraded at the same time, changes in evaluation results are hard to attribute.

1
Sources
50%
Confidence
Long-term
Relevance
10/16/2026
First Seen

Sources

Related Entities

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/1049502
API
curl https://kongchang.com/api/v1/knowledge/claims/1049502
MCP
get_claim(id=1049502)