mlsanityData to 7 Oct 2026

WATT: Weight Average Test-Time Adaptation of CLIP

arXiv 2406.13875 ↗First posted 19 Jun 2024cs.CVcited by 33

On the map: Multimodal LLMs / vision-language models → Prompt learning and adaptation of CLIP

Abstract (arXiv)

Vision-Language Models (VLMs) such as CLIP have yielded unprecedented performance for zero-shot image classification, yet their generalization capability may still be seriously challenged when confronted to domain shifts. In response, we present Weight Average Test-Time Adaptation (WATT) of CLIP, a pioneering approach facilitating full test-time adaptation (TTA) of this VLM. Our method employs a diverse set of templates for text prompts, augmenting the existing framework of CLIP. Predictions are utilized as pseudo labels for model updates, followed by weight averaging to consolidate the learned information globally. Furthermore, we introduce a text ensemble strategy, enhancing overall test performance by aggregating diverse textual cues. Our findings underscore the efficacy of WATT in enhancing performance across diverse datasets, including CIFAR-10-C, CIFAR-10.1, CIFAR-100-C, VisDA-C, and several other challenging datasets, effectively covering a wide range of domain shifts. Notably, these enhancements are achieved without necessitating additional model transformations or trainable modules. Moreover, compared to other Test-Time Adaptation methods, our approach can operate effectively with just a single image. Highlighting the potential of innovative test-time strategies, this research emphasizes their role in fortifying the adaptability of VLMs. The implementation is available at: urlhttps://github.com/Mehrdad-Noori/WATT.git.

Key facts. Cited by 33 papers in the map (data through 7 Oct 2026). Placed in Prompt learning and adaptation of CLIP, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 23 of its 33 references that are in the map. First posted on arXiv 19 Jun 2024.

Builds on

23 of 33 references.

Closest ancestors · 3

Same direction, newest first

  1. 2024Efficient Test-Time Adaptation of Vision-Language Models27 Mar
  2. 2023Diverse Data Augmentation with Diffusions for Effective Test-time Prompt Tuning11 Aug
  3. 2022Test-Time Prompt Tuning for Zero-Shot Generalization in Vision-Language Models15 Sep

Other references · 20

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2023Segment Anythingcited by 8,158
  3. 2021Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervisioncited by 3,848
  4. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  5. 2018Averaging Weights Leads to Wider Optima and Better Generalizationcited by 1,194
  6. 2017Deep Hashing Network for Unsupervised Domain Adaptationcited by 1,181
  7. 2017Deeper, Broader and Artier Domain Generalizationcited by 983
  8. 2018Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNscited by 585
  9. 2021SWAD: Domain Generalization by Seeking Flat Minimacited by 431
  10. 2021MEMO: Test Time Robustness via Adaptation and Augmentationcited by 376
  11. 2021AudioCLIP: Extending CLIP to Image, Text and Audiocited by 335
  12. 2020Evaluating Prediction-Time Batch Normalization for Robustness under Covariate Shiftcited by 235
8 more
  1. 2023CLIP-Driven Universal Model for Organ Segmentation and Tumor Detectioncited by 224
  2. 2022Test-Time Training with Masked Autoencoderscited by 208
  3. 2022Parameter-free Online Test-time Adaptationcited by 184
  4. 2022Frozen CLIP Models are Efficient Video Learnerscited by 181
  5. 2022TTTFlow: Unsupervised Test-Time Training with Normalizing Flowcited by 33
  6. 2023ClusT3: Information Invariant Test-Time Trainingcited by 21
  7. 2023Lookaround Optimizer: k steps around, 1 step averagecited by 9
  8. 2024NC-TTT: A Noise Contrastive Approach for Test-Time Trainingcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 3; 2025: 17; 2026: 13.

33 papers.

Sort:

By area:

  1. 2025Advances in Multimodal Adaptation and Generalization: From Traditional Approaches to Foundation Modelscited by 23
  2. 2025Ranked Entropy Minimization for Continual Test-Time Adaptationcited by 17
  3. 2025Realistic Test-Time Adaptation of Vision-Language Modelscited by 14
  4. 2024textttBATCLIP: Bimodal Online Test-Time Adaptation for CLIPcited by 11
  5. 2025Test-Time Adaptation of Vision-Language Models for Open-Vocabulary Semantic Segmentationcited by 10
  6. 2024A Simple and Efficient Baseline for Zero-Shot Generative Classificationcited by 7
  7. 2025LoRA-TTT: Low-Rank Test-Time Training for Vision-Language Modelscited by 7
  8. 2025Generalizing vision-language models to novel domains: A comprehensive surveycited by 6
  9. 2025Mint: A Simple Test-Time Adaptation of Vision-Language Models against Common Corruptionscited by 5
  10. 2026Test-Time Adaptation via Cache Personalization for Facial Expression Recognition in Videoscited by 5
  11. 2026Histopath-C: Towards Realistic Domain Shifts for Histopathology Vision-Language Adaptationcited by 4
  12. 2024Semantic Anchor Transport: Robust Test-Time Adaptation for Vision-Language Modelscited by 3
21 more
  1. 2025Test-time Correlation Alignmentcited by 3
  2. 2025Processing and acquisition traces in visual encoders: What does CLIP know about your camera?cited by 3
  3. 2025Purge-Gate: Backpropagation-Free Test-Time Adaptation for Point Clouds Classification via Token Purgingcited by 3
  4. 2025textttAVROBUSTBENCH: Benchmarking the Robustness of Audio-Visual Recognition Models at Test-Timecited by 2
  5. 2026CLIP-AUTT: Test-Time Personalization with Action Unit Prompting for Fine-Grained Video Emotion Recognitioncited by 2
  6. 2025FRET: Feature Redundancy Elimination for Test Time Adaptationcited by 1
  7. 2025Uniformity First: Uniformity-aware Test-time Adaptation of Vision-language Models against Image Corruptioncited by 1
  8. 2025Enhancing Monte Carlo Dropout Performance for Uncertainty Quantificationcited by 1
  9. 2025Adapt-As-You-Walk Through the Clouds: Training-Free Online Test-Time Adaptation of 3D Vision-Language Foundation Modelscited by 1
  10. 2026LQA: A Lightweight Quantized-Adaptive Framework for Vision-Language Models on the Edgecited by 1
  11. 2026ORION: ORthonormal Text Encoding for Universal VLM AdaptatIONcited by 1
  12. 2026Multi-Hypothesis Test-Time Adaptation to Mitigate Underspecificationcited by 1
  13. 2026Can Experts Adapt Without Training? On Test-Time Modality Generalization in MVLMscited by 1
  14. 2025TransForSeg: A Multitask Stereo ViT for Joint Stereo Segmentation and 3D Force Estimation in Catheterizationcited by 0
  15. 2025TRUST: Test-Time Refinement using Uncertainty-Guided SSM Traversescited by 0
  16. 2026Ramen: Robust Test-Time Adaptation of Vision-Language Models with Active Sample Selectioncited by 0
  17. 2026SHED: Style-Homogenized Embedding Alignment for Domain Generalizationcited by 0
  18. 2026Test-Time Adaptation via Dual Distillation for Videos Under Severe Distribution Shiftscited by 0
  19. 2026Test-Time Adaptation with Online Personalized Energy-Based Cache for Fine-Grained Video Expression Recognitioncited by 0
  20. 2026From Inference to Adaptation: A Unified Optimal Transport View of Vision Language Modelcited by 0
  21. 2026Balanced Prompt Adaptation against Entropy-Induced Collapse for Test-Time Binary Segmentationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024CLIPArTT: Adaptation of CLIP to New Domains at Test Time
  2. 2026What Drives Test-Time Adaptation for CLIP? A Controlled Empirical Study from an Update Perspective
  3. 2025The Illusion of Progress? A Critical Look at Test-Time Adaptation for Vision-Language Models
  4. 2025Ultra-Light Test-Time Adaptation for Vision--Language Models
  5. 2025ETTA: Efficient Test-Time Adaptation for Vision-Language Models through Dynamic Embedding Updates
  6. 2024A Lost Opportunity for Vision-Language Models: A Comparative Study of Online Test-Time Adaptation for Vision-Language Models
  7. 2025Bayesian Test-Time Adaptation for Vision-Language Models
  8. 2025Unleashing the Potential of All Test Samples: Mean-Shift Guided Test-Time Adaptation