Unverified50% confidenceFactExact time
The research team fine-tuned the Qwen3.5 series of small language models on filtered and rebalanced teacher-generated supervised data.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/9/2026
First Seen
Valid until: 1/7/2027
Sources
Related Entities
Related Claims
UnverifiedThe researchers validated the parity conclusion on the vision-language model Qwen2.5-VL-3B-Instruct for a single compression ratio using per-scheme learning rate tuning repeated with three random seeds.68% similarUnverifiedOn-policy distillation originates from the DAGGER-related paper and can be viewed as iterative SFT that incorporates teacher evaluation into a closed loop, with applications in GLM-4.5, Qwen3, and DeepSeek.67% similarUnverifiedIn the large language model domain, the industry has shifted to a synthetic data distillation approach: having teacher models produce input-output pairs in bulk and then performing standard supervised fine-tuning (SFT) on student models.67% similarVerifiedIn offline distillation, training data is generated once by the teacher model and fixed, with the student model learning from a static dataset.66% similarUnverifiedQwen3采用四阶段训练流程，在后期专门针对混合推理能力进行对齐微调66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/990329API
curl https://kongchang.com/api/v1/knowledge/claims/990329MCP
get_claim(id=990329)