Unverified50% confidenceBenchmarkExact time
In the e-commerce product classification task, the fine-tuned Qwen 3.5 2B model achieved an accuracy of 69.8%, surpassing the 68.2% of the un-fine-tuned GPT-6 Luna.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/18/2026
First Seen
Valid until: 2/16/2027
Sources
Small Model Fine-Tuning Upset: 2B/4B Parameters Beat GPT-6 Luna
youtubeNeuralNine11/18/2026
Related Entities
Related Claims
UnverifiedOn "continuation of previous operation" type utterances, Qwen3.6-35B-A3B achieved an accuracy of 86.3%, while the 4B and 8B models achieved only 70.4%.76% similarUnverifiedIn the experiment, Qwen 3.8 27B reached 88% recognition confidence for Breakout.75% similarUnverifiedOn the held-out compound recipe test set, the LoRA fine-tuned model caught 66%, while the base Qwen3.5-4B model caught 78%.75% similarUnverifiedIn the request type judgment task, the ordinary qwen3:8b and qwen3.5:4b models achieved an accuracy of about 82%.74% similarUnverifiedIn agentgateway's testing, qwen3.8 (27.3B, Q4_K_M quantization, 262K context) had a valid answer rate of 90.0% (54/60), an accuracy rate among answered questions of 98.1% (53/54), and an overall accuracy rate of 88.3% (53/60).73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1185316API
curl https://kongchang.com/api/v1/knowledge/claims/1185316MCP
get_claim(id=1185316)