Unverified50% confidenceFactExact time
The tests used Claude Sonnet 5.5 and Claude Opus 5.5 as answering models and Claude Haiku 5.5 as judge, producing 40 sets of results in total.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/19/2026
First Seen
Valid until: 1/17/2027
Sources
Related Entities
Related Claims
UnverifiedIn the tests, each Claude model (Haiku 4.5, Haiku 5.5, Sonnet 5.5, Opus 5.5) exhibited outcome-bias-induced score leakage on at least one case.79% similarUnverifiedClaude Sonnet 5.5 achieves Opus 5-level accuracy in document understanding, parsing, and search.73% similarUnverifiedIn agentic search tests, Sonnet 5 outperformed Sonnet 4.6 across all effort levels, but its overall performance was below Opus 4.8, placing it at approximately the Opus 4.7 level overall.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1065259API
curl https://kongchang.com/api/v1/knowledge/claims/1065259MCP
get_claim(id=1065259)