Unverified50% confidenceFactExact time
During training, MRL simultaneously optimizes loss functions at multiple dimensional levels, with the total loss being a weighted sum of losses at each granularity.
1
Sources
50%
Confidence
Long-term
Relevance
8/11/2026
First Seen
Sources
Embedding降维：Matryoshka套娃表征与PCA方法深度对比
Related Claims
Unverified常见的对比学习损失函数包括 NT-Xent Loss 和 Triplet Loss72% similarUnverifiedThe algorithm's loss function consists of three components: Focal Loss (positive-negative sample balancing), similarity loss (adjacent point consistency constraint), and shape loss (second-order smoothness constraint).69% similarUnverifiedThe core loss functions of contrastive learning (such as InfoNCE and NT-Xent) essentially pull the representations of similar samples closer together while pushing the representations of dissimilar samples further apart.67% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/729521API
curl https://kongchang.com/api/v1/knowledge/claims/729521MCP
get_claim(id=729521)