Unverified50% confidenceBenchmarkExact time

The fine-tuned Qwen3.5-4B model improved accuracy on BU Bench V1 from 15% to 53%.

1

Sources

50%

Confidence

Long-term

Relevance

9/7/2026

First Seen

Sources

Fine-Tuning a 4B Small Model: Browser Task Accuracy Soars from 22% to 63%

Related Entities

Qwen3.5-4Bmodel BU Bench V1concept

Related Claims

UnverifiedIn the Terminal Bench test, Qwen 3.6 27B scored approximately 60.7, while Muse Glimmer scored only 51.7, falling significantly behind.75% similar UnverifiedQwen 3.6 Max has an end-to-end accuracy of only 12.3% in neutral mode73% similar VerifiedQwen3.5's parameter activation rate is 4.3%, while DeepSeek V3's 671B model has an activation rate of 5.5%.72% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/869643

API

curl https://kongchang.com/api/v1/knowledge/claims/869643

MCP

get_claim(id=869643)