AI Paper Atlas

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

Branch, or Layer? Zeroth-Order Optimization for Continual Learning of Vision-Language Models

Ziwei Liu, Borui Kang, Wei Li, Hangjie Yuan, Yanbing Yang, Wenbin Li, Yifan Zhu, Tao Feng, Jun Luo

AAAI 2026 · Vision-language and multimodal › CLIP and zero-shot · approach: Test-time adaptation of VLMs · 2 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
replace FO with ZO in both branches (ablation)MoE4Adapter VLCL · CIFAR-100 Inc20Last accuracy: 80.47 -> 69.29Hurt
apply ZO only in language branch instead of both (ablation)MoE4Adapter VLCL · CIFAR-100 Inc20Last accuracy: 69.29 -> 80.94Improved
apply ZO in language branch vs vision branch (ablation)MoE4Adapter VLCL · TinyImageNet Inc20Last accuracy: 49.65 -> 49.14Mixed
layer-wise interleaved ZO (Hop-odd) in dual branch instead of all-layer ZO (ablation)MoE4Adapter VLCL · CIFAR-100 Inc20Last accuracy: 69.29 -> 81.64Improved
interleaved layer ZO (Hop-even) in vision branch vs full-FO baseline (ablation)MoE4Adapter VLCL · CIFAR-100 Inc20Last accuracy: 80.47 -> 82.41Improved
contiguous Prefix (six) ZO vs interleaved Hop-odd in dual branch (ablation)MoE4Adapter VLCL · CIFAR-100 Inc20Last accuracy: 81.64 -> 79.12Hurt
add MoZO (sign gradient + smaller vision perturbation) to Hop-even dual (ablation)Dual w/ Hop-even · TinyImageNet Inc20Avg accuracy: 60.53 -> 61.23Improved
add MoZO to Hop-odd dual on CIFAR Inc10 (ablation)Dual w/ Hop-odd · CIFAR-100 Inc10Avg accuracy: 86.98 -> 87.02No clear effect
add loss-validated conservative ZO instead of naive single-estimate ZO (ablation)Language w/ ZO · CIFAR-100 Inc10Last accuracy: 76.19 -> 76.74Improved
add gradient sign transform to naive ZO in language branch (ablation)Language w/ ZO* · CIFAR-100 Inc10Last accuracy: 76.19 -> 78.52Improved
apply ZO instead of FO in dual branch (memory)MoE4Adapter VLCL · GPU memorymemory: ~19.96GB -> ~2.17GBImproved

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