arXiv Machine Learning August 3, 2026 By Pouya Parsa, Raoof Zare Moayedi, Seongjin Choi

Visual Distribution Anchoring for Efficient Prompt Tuning

Read the original on arXiv Machine Learning →

arXiv:2607. 28967v1 Announce Type: cross Abstract: Prompt tuning adapts vision--language models with few trainable parameters, but existing approaches trade off efficiency and adaptation: static textual prompts can overfit source classes, image-conditioned prompts add per-instance computation, and multimodal tuning modifies the visual branch.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv Machine Learning.

arXiv AI

Jul 9

T2T-VICL: Cross-Task Visual In-Context Learning via Implicit Text-Driven VLMs

arXiv:2511. 16107v3 Announce Type: replace-cross Abstract: Visual in-context learning (VICL) solves visual tasks by conditioning on a few input-output demonstrations without any model training.

By Shao-Jun Xia, Huixin Zhang, Zhengzhong Tu

llms multimodal safety

arXiv Machine Learning

Jul 2

AdaBoosting Text Prompts for Vision-Language Models

arXiv:2607. 00684v1 Announce Type: new Abstract: The classification accuracy of pretrained Vision-Language Models (VLMs) relies on the quality of the text prompts.

By Seokhee Jin, Changhwan Sung, Sunung Mun, Hoyoung Kim, Jungseul Ok

llms multimodal benchmarks

arXiv AI

Jun 15

Learning What to Predict: Downstream-Guided Task Design for Continued Pretraining

arXiv:2601. 22108v2 Announce Type: replace-cross Abstract: Continued pretraining is optimized with fixed self-supervised tasks but selected by downstream performance, creating a coarse feedback loop in which practitioners evaluate checkpoints, change data mixtures or objectives, and restart runs, while individual updates remain blind to target capabilities.

By Shuqi Ke, Giulia Fanti

computer-vision

arXiv AI

Jul 22

Attributes Should Come from Images, Not Class Names: Distribution-Conditioned Attribute Selection for Vision-Language Models

arXiv:2607. 18695v1 Announce Type: cross Abstract: A popular route to interpretable zero-shot classification asks a large language model (LLM) to describe each class name and prompts CLIP with the resulting descriptors.

By Gautam Rajendrakumar Gare, Jia Shi, Zhiqiu Lin, Deepak Pathak, John Galeotti, Deva Ramanan

llms rag multimodal

arXiv Machine Learning

Jul 17

CARPRT: Class-Aware Zero-Shot Prompt Reweighting for Black-Box Vision-Language Models

arXiv:2607. 14125v1 Announce Type: new Abstract: Pre-trained vision-language models (VLMs) enable zero-shot image classification by computing the similarity score between an image and textual descriptions, typically formed by inserting a class label (e.

By Ruijiang Dong, Zesheng Ye, Jianzhong Qi, Lei Feng, Feng Liu, Gang Niu, Masashi Sugiyama

llms computer-vision multimodal benchmarks

arXiv AI

Jun 4

EvoPrompt: Guided Prompt Evolution for Vision-Language Models Adaptation

arXiv:2603. 09493v2 Announce Type: replace-cross Abstract: The adaptation of large-scale vision-language models (VLMs) to downstream tasks with limited labeled data remains a significant challenge.

By Enming Zhang, Jiayang Li, Yanlong Wang, Yanru Wu, Zhenyu Liu, Yang Li

llms rag fine-tuning multimodal benchmarks

Related stories

T2T-VICL: Cross-Task Visual In-Context Learning via Implicit Text-Driven VLMs

AdaBoosting Text Prompts for Vision-Language Models

Learning What to Predict: Downstream-Guided Task Design for Continued Pretraining

Attributes Should Come from Images, Not Class Names: Distribution-Conditioned Attribute Selection for Vision-Language Models

CARPRT: Class-Aware Zero-Shot Prompt Reweighting for Black-Box Vision-Language Models

EvoPrompt: Guided Prompt Evolution for Vision-Language Models Adaptation