[KongchangAI]
Unverified50% confidenceBenchmarkExact time

In the τ³-bench banking experiments, when Claude Haiku 5.5 drives both the solver and the evolver, skill gains reach as high as 21.5 percentage points.

1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/20/2026
First Seen
Valid until: 1/18/2027

Sources

EvoSkill v2: How a Stateful Evolver Lets AI Agents Rewrite Their Own Skills

Related Entities

Claude Haiku 5.5modelEvoSkill v2modelτ³-bench bankingconcept

Related Claims

UnverifiedHaiku 5.5 shows a notable capability leap between different reasoning intensity levels, with a considerable improvement from high to max/X-high, unlike Opus 5.5 or Sonnet.69% similar

Cite This Claim

Stable URI
https://kongchang.com/claim/1074286
API
curl https://kongchang.com/api/v1/knowledge/claims/1074286
MCP
get_claim(id=1074286)