Unverified50% confidenceFactExact time
Contrastive learning is the core paradigm for fine-tuning dense retrieval models, pulling a query and its positive document closer in vector space while pushing the query away from other documents in the batch (negatives).
1
Sources
50%
Confidence
Long-term
Relevance
10/26/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedInductive bias refers to the inherent tendency of a learning model to generalize on unseen data, arising from the constraints that model architecture, hypothesis space or prior knowledge place on solutions.76% similarUnverifiedEmbedding models convert queries and candidate documents into vector representations, thereby measuring relevance in a semantic space.76% similarUnverifiedModern large models learn from massive data and can search for strategies within the problem-solving space, no longer limited to single-formula retrieval.75% similarUnverifiedSPLADE 是一种学习型稀疏检索模型，通过神经网络为文档生成稀疏向量表示，可在倒排索引上高效运行75% similarUnverifiedKnowledge base retrieval finds the most relevant document fragments by computing the cosine similarity between the query vector and document vectors, then passes them to a large model to generate an answer.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1098464API
curl https://kongchang.com/api/v1/knowledge/claims/1098464MCP
get_claim(id=1098464)