AI Paper Atlas

Research map / Vision-language and multimodal / CLIP and zero-shot

Improving CLIP Adaptation by Breaking Tail Alignment for Source-Free Cross-Domain Few-Shot Learning

Shuai Yi, Yixiong Zou, Yuhua Li, Ruixuan Li

ICML 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Open-vocabulary alignment from VLMs · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add push-away of tail tokens (ablation)CLIP-LoRA · CDFSL 5-way 5-shot avgaccuracy: 65.99 -> 67.45Improved
add head-token enhancement on top of tail repulsion (ablation)Push-away tail · CDFSL 5-way 5-shot avgaccuracy: 67.45 -> 68.53Improved
use direct feature modulation instead of loss constraint (ablation)ATHA · CDFSL 5-way 5-shot avgaccuracy: 66.76 -> 68.53Improved
use MaxSim/MinSim text instead of weighted average text (ablation)ATHA · CDFSL 5-way 5-shot avgaccuracy: 67.73 -> 68.53Improved
make alpha, beta learnable per layer instead of fixed (ablation)ATHA · CDFSL 5-way 5-shot avgaccuracy: 68.26 -> 68.53Improved
add PACL alignment loss to LoRACLIP-LoRA · CDFSL 5-way 5-shot avgaccuracy: 65.99 -> 65.85Hurt
add SPARC alignment to LoRACLIP-LoRA · CDFSL 5-way 5-shot avgaccuracy: 65.99 -> 65.55Hurt
replace cosine similarity with Euclidean distance for token selection (ablation)ATHA · CDFSL 5-way 5-shot avgaccuracy: 68.53 -> 68.39No clear effect
add ATHA to CLIP ViT-L/14 LoRACLIP ViT-L/14 · CDFSL 5-way 1-shot avgaccuracy: 54.90 -> 56.90Improved
add ATHA to SigLIP LoRASigLIP · CDFSL 5-way 1-shot avgaccuracy: 49.83 -> 53.11Improved
add ATHA to PEcore LoRAPEcore · CDFSL 5-way 1-shot avgaccuracy: 56.61 -> 58.29Improved
add ATHA to CoOpCoOp · CDFSL 5-way 5-shot avgaccuracy: 60.29 -> 62.03Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy