Pith. sign in

REVIEW 3 major objections 5 minor 3 cited by

TextRefiner: Internal Visual Feature as Efficient Refiner for Vision-Language Models Prompt Tuning

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Patch tokens refine text prompts, lifting CoOp by 5.28 points

desk verdict A promising plug-and-play idea for refining VLM text prompts, but the headline CoOp gain rides on an unfair epoch comparison. read the letter →

arxiv 2412.08176 v1 pith:HGYCSCCM submitted 2024-12-11 cs.CV cs.MM

Jingjing Xie , Yuxin Zhang , Jun Peng , Zhaohong Huang , Liujuan Cao This is my paper · ORCID
classification cs.CVcs.MM
keywords TextRefinerprompttuningvision-languagemodelslocalcachefine-grainedvisualfeaturesCLIPbase-to-novelgeneralizationinferenceefficiency
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that the fine-grained visual details needed to distinguish similar classes are already present in CLIP's image-branch patch tokens, and that a plug-and-play module can extract them to refine text prompts. If this is right, prompt-tuning methods can match or beat LLM-based description generation while keeping inference fast and needing no external knowledge. The authors show TextRefiner raising CoOp's harmonic mean of base and novel accuracy from 71.66% to 76.94% over 11 benchmarks, and PromptKD with TextRefiner reaching an 82.33 harmonic mean at 12,793 FPS.

What carries the argument

The central object is the local cache module: a fixed-size storage $A \in \mathbb{R}^{M \times d}$ whose entries are updated by cosine-similarity-weighted aggregation of patch tokens, so each entry condenses a recurring visual attribute. Feature aggregation retrieves attribute embeddings by softmax-matching each text embedding to cache entries and fusing them through a residual linear layer (Eq. 7). Feature alignment, a 2-layer MLP (Eq. 8), maps patch tokens into the text embedding space to close the modality gap, and two auxiliary losses, semantic and regularization, supervise the alignment.

What would settle it

A direct test is to take a trained TextRefiner and measure whether each cache entry's nearest patch tokens are class-consistent on a held-out split; if entries do not separate classes better than random, the mechanism is not doing what is claimed. A cheaper check is to drop the feature alignment MLP entirely and see whether the cache alone still improves CoOp.

Watch

Extended reading notes

Core claim

TextRefiner establishes that internal visual features, specifically the local patch tokens of a ViT image encoder, can serve as a source of class-level visual attributes for refining text embeddings in CLIP prompt tuning. The central claim is that this internal-knowledge route is as effective as external LLM descriptions while being nearly inference-free. During tuning, patch tokens are clustered into a local cache whose entries act as attribute priors; at inference these entries are matched to text embeddings and fused through a linear layer, with an MLP aligning local visual features into text space. The paper reports consistent gains across CoOp, MaPLe, and PromptKD, including large novel-class improvements.

Load-bearing premise

The whole method rests on the assumption that CLIP image-branch patch tokens contain class-discriminative fine-grained attributes that can be clustered into a shared cache and mapped into text space by a shallow MLP; if that assumption fails, the cache stores noise and the refinement cannot help novel classes.

Editorial extensions

If this is right

  • Any existing prompt-tuning method that keeps CLIP's text encoder can be upgraded by plugging in TextRefiner, since the cache operates on text output embeddings rather than requiring retraining the backbone.
  • Novel-class generalization improves without instance-wise conditioning: TextRefiner lifts CoOp's novel accuracy from 63.22% to 74.32%, exceeding CoCoOp's 71.69%.
  • Cross-domain robustness improves on ImageNet variants, with CoOp plus TextRefiner outperforming plain CoOp on ImageNet-V2, ImageNet-Sketch, ImageNet-A, and ImageNet-R.
  • The approach competes with LLM-based prompt refinement at far lower inference cost: PromptKD with TextRefiner runs at 12,793.26 FPS versus LLaMP's 1,473.46 FPS.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension the paper does not test is whether the same local cache can be reused across datasets as a transferable visual-attribute vocabulary; if entries are class-agnostic texture or part detectors, a cache trained on one dataset should refine prompts on a held-out dataset.
  • The cache size $M$ is a sensitivity point: the paper's ablation shows accuracy peaks then falls as $M$ grows, suggesting that beyond a threshold the clustering fragments whole attributes; this could be turned into an automatic criterion for choosing $M$ from validation-set harmonic mean.
  • The feature alignment MLP may be doing much of the reported work, and the paper does not isolate whether gains come from the cache content or from the learned projection; replacing Eq. 8 with a fixed linear projection would separate these effects.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper proposes TextRefiner, a plug-and-play module that refines the text prompts of vision-language model (VLM) prompt-tuning methods using internal visual features. A local cache stores clusters of image patch tokens, a feature aggregation module fuses the cached visual descriptions into class text embeddings, and a feature alignment module (a 2-layer MLP) maps local tokens into the text embedding space. Training adds a semantic loss and a regularization loss to the standard contrastive loss. The authors evaluate on base-to-novel generalization across 11 datasets, cross-domain generalization on ImageNet variants, and inference efficiency, reporting consistent gains when TextRefiner is attached to CoOp, MaPLe, and PromptKD, with PromptKD w/TextRefiner achieving the best average harmonic mean.

Significance. If the reported results are reproducible, TextRefiner is a useful contribution: it improves the base-to-novel generalization of several prompt-tuning baselines without external LLM knowledge or heavy inference overhead, and it is orthogonal to the underlying prompt-tuning method. The paper ships code, reports a large-scale evaluation, and explicitly adopts the standard base-to-novel protocol, which is a genuine generalization check. The method's motivation (using internal local visual features to refine text prompts) is plausible, and the efficiency numbers are striking. However, the headline CoOp comparison is marred by an unequal training schedule, and several method details need clarification before the central claim can be accepted.

major comments (3)
  1. [Implementation details / Table 1] The strongest claimed result, "improves CoOp from 71.66% to 76.94%," is confounded by training budget. The Implementation details section states that all settings match the original papers "except for CoOp, where the number of epochs was reduced from 200 to 10." Thus Table 1 compares a 200-epoch CoOp baseline with a 10-epoch CoOp+TextRefiner variant. Since fewer epochs is itself a strong regularizer in few-shot prompt tuning, the observed shift (base accuracy dropping from 82.69 to 79.74 while novel accuracy rises from 63.22 to 74.32) is exactly what one would expect from early stopping, not necessarily from the proposed cache, aggregation, or alignment modules. The authors should report a 10-epoch CoOp baseline, or a 200-epoch CoOp+TextRefiner variant, or otherwise isolate the effect of TextRefiner from the schedule change.
  2. [Table 3 / Ablations] The ablation study in Table 3 uses the original CoOp row as the empty-component baseline, but the TextRefiner rows are trained for 10 epochs. Therefore the incremental gains attributed to the local cache, semantic loss, and regularization loss are not cleanly isolated from the change in training epochs. A matched-schedule baseline is needed before the component contributions can be interpreted. In addition, none of the tables report multiple seeds or error bars; given the small margins on some datasets (e.g., Caltech101 and OxfordPets rows), variance estimates are important for assessing whether the gains are significant.
  3. [Local Cache / Eqs. (2)-(4)] The text repeatedly states that the local cache stores fine-grained information "in a class-wise manner," but the update rule in Eqs. (2)-(4) is a global cache: every local token from every class is assigned to one of M shared entries, with no class label or class-conditional term in the cosine-similarity assignment or the momentum update. The subsequent querying with class text embeddings (Eqs. (5)-(6)) does not make the stored entries class-specific. The method may work perfectly well with a shared attribute cache, but the paper should either clarify what "class-wise" means, or replace the description with the actual global-cache mechanism, since the current text misdescribes the algorithm.
minor comments (5)
  1. [Implementation details / Method notation] There is a notation mismatch: the Implementation details mention "momentum α and the fusion factor β," but the method defines the momentum coefficient as γ in Eq. (4) and the aggregation coefficient as α in Eq. (7). Please align the notation.
  2. [Feature Alignment / Eq. (10)] The semantic loss in Eq. (10) refers to "top-k transformed local features with attention scores," but no attention scores are defined in the method. Please specify how the top-k local features are selected and where the attention scores come from.
  3. [Method / local tokens source] The paper does not state which layer of the ViT image encoder provides the local tokens V. Since the whole method depends on these tokens carrying fine-grained visual concepts, please specify the layer and, ideally, provide an analysis of how the choice of layer affects results.
  4. [Abstract] There is a typo: "relesed" should be "released." Similar typos appear elsewhere (e.g., "the transfer" in the preliminaries), so a proofreading pass is recommended.
  5. [Efficiency comparison / Figure 3] The FPS comparison in Figure 3 is reported only for ImageNet and without a precise protocol description (batch size, input resolution, hardware details beyond "one single A800 GPU"). Please state the measurement protocol so the efficiency claim is reproducible.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: TextRefiner is an empirical plug-in evaluated on held-out classes; the CoOp epoch mismatch is a protocol concern, not a circular derivation.

full rationale

No circularity found. TextRefiner's claimed gains are empirical: the method adds a local cache, feature aggregation, feature alignment, and semantic/regularization losses to existing prompt-tuning baselines, and is measured on held-out novel classes and cross-domain sets. The refined text embedding is computed by Eqs. 6-8 from cache entries that are updated by image-branch patch tokens (Eq. 4); no target metric or label is defined in terms of these fitted quantities. The only notable protocol issue is the Implementation details statement that 'the number of epochs was reduced from 200 to 10' for CoOp, which confounds the headline CoOp comparison; this is a fairness/ablation-control concern, not circularity, because the evaluation remains external and the method's components do not reduce to its inputs by construction. Hyperparameter choices (lambda1, lambda2, alpha, M) are ordinary ablation-selected settings, not fitted predictions. No load-bearing self-citation chain or imported uniqueness theorem is present; cited evidence for patch-token interpretability is external and non-circular.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

All hyperparameters in the method are chosen by hand or through ablations; none are derived from first principles. The paper introduces no new physical entities; its inventions are architectural (cache, MLP, losses). The key untested premise is that CLIP patch tokens carry class-discriminative semantics that a shallow MLP can align with text space.

free parameters (6)
  • M (cache size) = not stated in paper; ablation in Fig. 4 over 5-40
    Number of cache entries for summarizing local tokens; chosen by hand via ablation; exact default absent from text.
  • gamma (cache momentum coefficient in Eq. 4) = 0.8 (reported as 'momentum alpha' in Implementation details)
    Controls how much new local tokens update each cache entry; hand-set and not derived.
  • alpha (feature aggregation coefficient in Eq. 7) = 0.2 (reported as 'fusion factor beta' in Implementation details)
    Blends local and global text embeddings; tuned via ablation in Fig. 5.
  • lambda1 (semantic loss weight) = 0.02
    Chosen by 'empirical observation'; tunes alignment of local features to text embeddings.
  • lambda2 (regularization loss weight) = 20
    Chosen by 'empirical observation'; prevents overfitting on few-shot data.
  • k (top-k local features in semantic loss) = not specified in paper
    Eq. 10 uses top-k patch tokens by attention, but k is never stated; code would be needed.
assumptions (4)
  • domain assumption Patch tokens of the CLIP vision transformer encode fine-grained visual concepts (edges, textures, object parts).
    Cited to Ghiasi et al. and Zeiler & Fergus; assumed to hold for CLIP ViT-B/16 and to be class-discriminative enough for prompt refinement. Invoked in 'Local Cache'.
  • ad hoc to paper A 2-layer MLP can transform local visual tokens into the text embedding space sufficiently to reduce the modality gap.
    The MLP in Eq. 8 is introduced specifically for this method; no independent evidence beyond the empirical results.
  • ad hoc to paper Momentum-updated cosine-similarity assignment (Eqs. 2-4) yields stable cache entries that represent recurring visual attributes rather than instance noise.
    No convergence or stability analysis is provided; relies on heuristic soft-assignment updates.
  • domain assumption Base-class cache generalizes to novel classes in base-to-novel evaluation.
    The central generalization test assumes cache entries learned on base classes transfer to unseen classes; the paper reports results for this but no mechanism.

how reviews work

0 comments
Cite this review

Pith. "Pith review of TextRefiner: Internal Visual Feature as Efficient Refiner for Vision-Language Models Prompt Tuning." pith.science (2026). https://pith.science/paper/HGYCSCCM

@misc{pith2026241208176,
  author       = {Pith},
  title        = {Pith review of: TextRefiner: Internal Visual Feature as Efficient Refiner for Vision-Language Models Prompt Tuning},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HGYCSCCM}},
  note         = {Machine review of arXiv:2412.08176}
}
read the original abstract

Despite the efficiency of prompt learning in transferring vision-language models (VLMs) to downstream tasks, existing methods mainly learn the prompts in a coarse-grained manner where the learned prompt vectors are shared across all categories. Consequently, the tailored prompts often fail to discern class-specific visual concepts, thereby hindering the transferred performance for classes that share similar or complex visual attributes. Recent advances mitigate this challenge by leveraging external knowledge from Large Language Models (LLMs) to furnish class descriptions, yet incurring notable inference costs. In this paper, we introduce TextRefiner, a plug-and-play method to refine the text prompts of existing methods by leveraging the internal knowledge of VLMs. Particularly, TextRefiner builds a novel local cache module to encapsulate fine-grained visual concepts derivedfrom local tokens within the image branch. By aggregating and aligning the cached visual descriptions with the original output of the text branch, TextRefiner can efficiently refine and enrich the learned prompts from existing methods without relying on any external expertise. For example, it improves the performance of CoOp from 71.66 % to 76.94 % on 11 benchmarks, surpassing CoCoOp which introduces instance-wise features for text prompts. Equipped with TextRefiner, PromptKD achieves state-of-the-art performance and is efficient in inference. Our code is relesed at https://github.com/xjjxmu/TextRefiner

Figures

Figures reproduced from arXiv: 2412.08176 by the authors.

Figure 1
Figure 1. Comparison of different paradigms for enriching [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. The framework of TextRefiner, which is composed of local cache, feature aggregation and feature alignment. Here, [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Comparison of inference efficiency among existing [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Ablation study on M in A. 1 λ1 20 30 40 50 60 Base 76.33 76.45 76.52 76.84 76.60 Novel 70.14 70.04 70.28 70.54 70.27 HM 73.10 73.10 73.27 73.56 73.30 λ2 5 10 15 20 25 Base 76.65 76.70 76.67 76.84 76.65 Novel 70.49 70.51 70.32 70.54 70.34 HM 73.44 73.47 73.36 73.56 73.3…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. NEARL: Interacted Query Adaptation with Orthogonal Regularization for Medical Vision-Language Understanding

    cs.CV 2025-08 conditional novelty 6.0 of 10

    NEARL-CLIP reports top accuracy on three medical image classification benchmarks by adding bidirectional cross-modal attention and orthogonalized adapters to CLIP with only 1.46M parameters.

  2. Spotlighter: Revisiting Prompt Tuning from a Representative Mining View

    cs.CV 2025-08 reject novelty 5.0 of 10

    Spotlighter improves CLIP prompt tuning by selecting top-k visual tokens via a prototype-guided activation score, reporting higher accuracy and faster inference with supposedly only 21 extra parameters.

  3. Cross-Domain Few-Shot Learning via Multi-View Collaborative Optimization with Vision-Language Models

    cs.CV 2025-08 conditional novelty 4.0 of 10

    CoMuCo fine-tunes CLIP with two complementary expert branches and Laplace/Jeffreys regularizers, achieving state-of-the-art few-shot accuracy on a newly collected cross-domain benchmark.

Reference graph

Works this paper leans on

53 extracted references · 27 canonical work pages · cited by 3 Pith papers

  1. [1]

    Bossard, L.; Guillaumin, M.; and Gool, L. V. 2014. Food-101--mining discriminative components with random forests. In ECCV, 446--461. Springer

    work page 2014

  2. [2]

    Chen, T.; Kornblith, S.; Norouzi, M.; and Hinton, G. 2020. A simple framework for contrastive learning of visual representations. In International conference on machine learning, 1597--1607. PMLR

    2020

  3. [3]

    Cimpoi, M.; Maji, S.; Kokkinos, I.; Mohamed, S.; and Vedaldi, A. 2014. Describing textures in the wild. In CVPR, 3606--3613

    2014

  4. [4]

    Deng, J.; Dong, W.; Socher, R.; Li, L.-J.; Li, K.; and Fei-Fei, L. 2009. Imagenet: A large-scale hierarchical image database. In CVPR, 248--255. Ieee

    2009

  5. [5]

    Dosovitskiy, A.; Beyer, L.; Kolesnikov, A.; Weissenborn, D.; Zhai, X.; Unterthiner, T.; Dehghani, M.; Minderer, M.; Heigold, G.; Gelly, S.; Uszkoreit, J.; and Houlsby, N. 2021. An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale. In International Conference on Learning Representations, (ICLR)

    work page 2021

  6. [6]

    Fei-Fei, L.; Fergus, R.; and Perona, P. 2004. Learning generative visual models from few training examples: An incremental bayesian approach tested on 101 object categories. In CVPR workshop, 178--178. IEEE

    work page 2004

  7. [7]

    Gao, J.; Ruan, J.; Xiang, S.; Yu, Z.; Ji, K.; Xie, M.; Liu, T.; and Fu, Y. 2024 a . LAMM: Label Alignment for Multi-Modal Prompt Learning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 1815--1823

    work page 2024

  8. [8]

    Gao, P.; Geng, S.; Zhang, R.; Ma, T.; Fang, R.; Zhang, Y.; Li, H.; and Qiao, Y. 2024 b . Clip-adapter: Better vision-language models with feature adapters. International Journal of Computer Vision, 132(2): 581--595

    2024

Show all 53 references
  1. [9]

    G.; and Goldstein, T

    Ghiasi, A.; Kazemi, H.; Borgnia, E.; Reich, S.; Shu, M.; Goldblum, M.; Wilson, A. G.; and Goldstein, T. 2022. What do vision transformers learn? a visual exploration. arXiv preprint arXiv:2212.06727

    2022 arXiv

  2. [10]

    Gu, X.; Lin, T.-Y.; Kuo, W.; and Cui, Y. 2022. Open-vocabulary Object Detection via Vision and Language Knowledge Distillation. In International Conference on Learning Representations

    2022

  3. [11]

    He, K.; Zhang, X.; Ren, S.; and Sun, J. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 770--778

    2016

  4. [12]

    Helber, P.; Bischke, B.; Dengel, A.; and Borth, D. 2019. Eurosat: A novel dataset and deep learning benchmark for land use and land cover classification. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 12(7): 2217--2226

    2019

  5. [13]

    Hendrycks, D.; Basart, S.; Mu, N.; Kadavath, S.; Wang, F.; Dorundo, E.; Desai, R.; Zhu, T.; Parajuli, S.; Guo, M.; et al. 2021 a . The many faces of robustness: A critical analysis of out-of-distribution generalization. In Proceedings of the IEEE/CVF international conference o...

    2021

  6. [14]

    Hendrycks, D.; Zhao, K.; Basart, S.; Steinhardt, J.; and Song, D. 2021 b . Natural adversarial examples. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 15262--15271

    2021

  7. [15]

    Jia, C.; Yang, Y.; Xia, Y.; Chen, Y.-T.; Parekh, Z.; Pham, H.; Le, Q.; Sung, Y.-H.; Li, Z.; and Duerig, T. 2021. Scaling up visual and vision-language representation learning with noisy text supervision. In International conference on machine learning, 4904--4916. PMLR

    2021

  8. [16]

    U.; Rasheed, H.; Maaz, M.; Khan, S.; and Khan, F

    Khattak, M. U.; Rasheed, H.; Maaz, M.; Khan, S.; and Khan, F. S. 2023 a . Maple: Multi-modal prompt learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 19113--19122

    2023

  9. [17]

    U.; Wasim, S

    Khattak, M. U.; Wasim, S. T.; Naseer, M.; Khan, S.; Yang, M.-H.; and Khan, F. S. 2023 b . Self-regulating prompts: Foundational model adaptation without forgetting. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 15190--15200

    2023

  10. [18]

    Kim, S.; Nam, J.; and Ko, B. C. 2022 a . V i T - N e T : Interpretable Vision Transformers with Neural Tree Decoder. In International Conference on Machine Learning (ICLR), volume 162, 11162--11172

    2022

  11. [19]

    Kim, S.; Nam, J.; and Ko, B. C. 2022 b . Vit-net: Interpretable vision transformers with neural tree decoder. In International conference on machine learning, 11162--11172. PMLR

    2022

  12. [20]

    Krause, J.; Stark, M.; Deng, J.; and Fei-Fei, L. 2013. 3d object representations for fine-grained categorization. In ICCV workshops, 554--561

    2013

  13. [21]

    Kunananthaseelan, N.; Zhang, J.; and Harandi, M. 2024. LaViP:Language-Grounded Visual Prompts

    2024

  14. [22]

    Q.; Belongie, S.; Koltun, V.; and Ranftl, R

    Li, B.; Weinberger, K. Q.; Belongie, S.; Koltun, V.; and Ranftl, R. 2022. Language-driven Semantic Segmentation. In International Conference on Learning Representations

    2022

  15. [23]

    Li, X.; Lian, D.; Lu, Z.; Bai, J.; Chen, Z.; and Wang, X. 2024 a . Graphadapter: Tuning vision-language models with dual knowledge graph. Advances in Neural Information Processing Systems, 36

    2024

  16. [24]

    Li, Z.; Li, X.; Fu, X.; Zhang, X.; Wang, W.; Chen, S.; and Yang, J. 2024 b . Promptkd: Unsupervised prompt distillation for vision-language models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 26617--26626

    2024

  17. [25]

    Liu, Z.; Mao, H.; Wu, C.-Y.; Feichtenhofer, C.; Darrell, T.; and Xie, S. 2022. A convnet for the 2020s. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 11976--11986

    2022

  18. [26]

    Ma, J.; Bai, Y.; Zhong, B.; Zhang, W.; Yao, T.; and Mei, T. 2023. Visualizing and understanding patch interactions in vision transformer. IEEE Transactions on Neural Networks and Learning Systems

    2023

  19. [27]

    Maji, S.; Rahtu, E.; Kannala, J.; Blaschko, M.; and Vedaldi, A. 2013. Fine-grained visual classification of aircraft. arXiv preprint arXiv:1306.5151

    2013 arXiv

  20. [28]

    Nilsback, M.-E.; and Zisserman, A. 2008. Automated flower classification over a large number of classes. In 2008 Sixth Indian Conference on Computer Vision, Graphics & Image Processing, 722--729. IEEE

    2008

  21. [29]

    M.; Vedaldi, A.; Zisserman, A.; and Jawahar, C

    Parkhi, O. M.; Vedaldi, A.; Zisserman, A.; and Jawahar, C. 2012. Cats and dogs. In CVPR, 3498--3505. IEEE

    2012

  22. [30]

    W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; et al

    Radford, A.; Kim, J. W.; Hallacy, C.; Ramesh, A.; Goh, G.; Agarwal, S.; Sastry, G.; Askell, A.; Mishkin, P.; Clark, J.; et al. 2021. Learning transferable visual models from natural language supervision. In International Conference on Machine Learning (ICML), 8748--8763. PMLR

    2021

  23. [31]

    Recht, B.; Roelofs, R.; Schmidt, L.; and Shankar, V. 2019. Do imagenet classifiers generalize to imagenet? In International conference on machine learning, 5389--5400. PMLR

    2019

  24. [32]

    Roy, S.; and Etemad, A. 2024. Consistency-guided Prompt Learning for Vision-Language Models. In International Conference on Learning Representations, (ICLR)

    2024

  25. [33]

    R.; Cogswell, M.; Das, A.; Vedantam, R.; Parikh, D.; and Batra, D

    Selvaraju, R. R.; Cogswell, M.; Das, A.; Vedantam, R.; Parikh, D.; and Batra, D. 2020. Grad-CAM: visual explanations from deep networks via gradient-based localization. International Journal of Computer Vision (IJCV), 128: 336--359

    2020

  26. [34]

    E.; Keutzer, K.; and Darrell, T

    Shen, S.; Yang, S.; Zhang, T.; Zhai, B.; Gonzalez, J. E.; Keutzer, K.; and Darrell, T. 2024. Multitask vision-language prompt tuning. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 5656--5667

    2024

  27. [35]

    R.; and Shah, M

    Soomro, K.; Zamir, A. R.; and Shah, M. 2012. UCF101: A dataset of 101 human actions classes from videos in the wild. arXiv preprint arXiv:1212.0402

    2012 arXiv

  28. [36]

    Tian, X.; Zou, S.; Yang, Z.; and Zhang, J. 2024. ArGue: Attribute-Guided Prompt Tuning for Vision-Language Models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 28578--28587

    2024

  29. [37]

    Touvron, H.; Lavril, T.; Izacard, G.; Martinet, X.; Lachaux, M.-A.; Lacroix, T.; Rozi \`e re, B.; Goyal, N.; Hambro, E.; Azhar, F.; et al. 2023 a . Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971

    2023 arXiv

  30. [38]

    Touvron, H.; Martin, L.; Stone, K.; Albert, P.; Almahairi, A.; Babaei, Y.; Bashlykov, N.; Batra, S.; Bhargava, P.; Bhosale, S.; et al. 2023 b . Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288

    2023 arXiv

  31. [39]

    Wang, H.; Ge, S.; Lipton, Z.; and Xing, E. P. 2019. Learning robust global representations by penalizing local predictive power. Advances in Neural Information Processing Systems, 32

    2019

  32. [40]

    Wu, X.; Zhu, F.; Zhao, R.; and Li, H. 2023. Cora: Adapting clip for open-vocabulary detection with region prompting and anchor pre-matching. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 7031--7040

    2023

  33. [41]

    A.; Hays, J.; Torralba, A.; and Oliva, A

    Xiao, J.; Ehinger, K. A.; Hays, J.; Torralba, A.; and Oliva, A. 2016. Sun database: Exploring a large collection of scene categories. IJCV, 119(1): 3--22

    2016

  34. [42]

    Xin, Y.; Du, J.; Wang, Q.; Yan, K.; and Ding, S. 2024. Mmap: Multi-modal alignment prompt for cross-domain multi-task learning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 16076--16084

    2024

  35. [43]

    Xing, Y.; Wu, Q.; Cheng, D.; Zhang, S.; Liang, G.; Wang, P.; and Zhang, Y. 2023. Dual modality prompt tuning for vision-language pre-trained model. IEEE Transactions on Multimedia

    2023

  36. [44]

    Yu, T.; Lu, Z.; Jin, X.; Chen, Z.; and Wang, X. 2023. Task residual for tuning vision-language models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 10899--10909

    2023

  37. [45]

    D.; and Fergus, R

    Zeiler, M. D.; and Fergus, R. 2014. Visualizing and understanding convolutional networks. In European Conference on Computer Vision (ECCV), 818--833. Springer

    2014

  38. [46]

    Zhai, X.; Mustafa, B.; Kolesnikov, A.; and Beyer, L. 2023. Sigmoid loss for language image pre-training. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 11975--11986

    2023

  39. [47]

    Zhang, R.; Zeng, Z.; Guo, Z.; and Li, Y. 2022. Can language understand depth? In Proceedings of the 30th ACM International Conference on Multimedia, 6868--6874

    2022

  40. [48]

    Zheng, Z.; Wei, J.; Hu, X.; Zhu, H.; and Nevatia, R. 2024. Large Language Models are Good Prompt Learners for Low-Shot Image Classification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 28453--28462

    2024

  41. [49]

    C.; and Liu, Z

    Zhou, K.; Yang, J.; Loy, C. C.; and Liu, Z. 2022 a . Conditional prompt learning for vision-language models. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 16816--16825

    2022

  42. [50]

    C.; and Liu, Z

    Zhou, K.; Yang, J.; Loy, C. C.; and Liu, Z. 2022 b . Learning to prompt for vision-language models. International Journal of Computer Vision (IJCV), 130(9): 2337--2348

    2022

  43. [51]

    Zhu, B.; Niu, Y.; Han, Y.; Wu, Y.; and Zhang, H. 2023. Prompt-aligned gradient for prompt tuning. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 15659--15669

    2023

  44. [52]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all...

  45. [53]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.