Unverified50% confidenceSolutionExact time
IDM uses a fixed item-to-code mapping to combine the student model's probabilities at each code position into a joint item probability, thereby reconstructing a catalog-level item distribution and aligning it with the teacher model's item distribution.
1
Sources
50%
Confidence
Long-term
Relevance
10/26/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedKnowledge distillation transfers the full probability distribution information output by the teacher model via soft labels, producing an implicit data augmentation effect.68% similarUnverifiedThe teacher model outputs a probability distribution in item space, while the student model makes parallel predictions in code space; the two spaces are inconsistent and cannot be directly aligned.66% similarUnverifiedIn traditional white-box distillation, the student model can directly access the complete probability distribution (logits) of the teacher model's output layer, whereas commercial large language model APIs typically return only the final generated text or top-k token probabilities.66% similarUnverifiedIDM is better at improving recall, indicating that reconstructing the item distribution helps the model cover more relevant items; CMM performs better on ranking quality (NDCG), indicating that code marginal matching can optimize ranking more finely.66% similarVerifiedIn offline distillation, training data is generated once by the teacher model and fixed, with the student model learning from a static dataset.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1098102API
curl https://kongchang.com/api/v1/knowledge/claims/1098102MCP
get_claim(id=1098102)