Unverified50% confidenceFactExact time
Training only the key heads' query and key projections outperforms layer-matched-head and attention-matched-head control groups and transfers to HPatches and BLINK.
1
Sources
50%
Confidence
Long-term
Relevance
11/17/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedPixel-level contrastive learning treats pixels belonging to the same keypoint as positive samples and pixels of different keypoints as negative samples, and introduces learnable keypoint prototypes as anchors.68% similarUnverifiedImproving cross-image comparison ability does not necessarily require large-scale retraining; localizing the attention heads that carry the function and applying minimal targeted interventions may yield significant gains, offering a more economical approach than full fine-tuning for scenarios such as image-text comparison, document comparison, and visual grounding.67% similarUnverifiedThe attention mechanism assigns learnable weights to each channel or each position of the feature map, achieving soft feature selection, with the weights determined in a data-driven rather than manually specified manner.66% similarUnverifiedAn induction head is usually implemented by two attention heads working together: the previous-token head marks each token's predecessor, and the induction head uses that marker to retrieve the successor.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1179122API
curl https://kongchang.com/api/v1/knowledge/claims/1179122MCP
get_claim(id=1179122)