Unverified50% confidenceOpinionExact time
The MAEM method still currently uses Qwen3.6-27B as its main experimental subject, and its generalization ability across models of different architectures remains to be further verified.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/6/2026
First Seen
Valid until: 2/4/2027
Sources
Related Entities
Related Claims
UnverifiedTev1-4B-experimental is built on top of the Qwen3.5 4B base model.74% similarUnverifiedThe Agent's default self-hosted configuration uses the Qwen3.6-35B-A3B model with 4-bit quantization deployed via sglang.71% similarUnverifiedQwen3.6-27B uses a Dense architecture rather than a MoE (Mixture of Experts) architecture.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1142265API
curl https://kongchang.com/api/v1/knowledge/claims/1142265MCP
get_claim(id=1142265)