[KongchangAI]
Unverified50% confidenceOpinionExact time

The study infers that model combinations significantly affect judging results, and that LLM judges may have self-preference or style blind spots.

1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/11/2026
First Seen
Valid until: 2/9/2027

Sources

Related Entities

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/1160921
API
curl https://kongchang.com/api/v1/knowledge/claims/1160921
MCP
get_claim(id=1160921)