Pith. sign in

REVIEW 5 major objections 6 minor 41 references

Selective Visual Prompting in Vision Mamba

T0 review · 5 major / 6 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 Token-wise selective prompts adapt frozen Vision Mamba better than prefix prompts.

desk verdict SVP is a solid empirical win for prompting Vision Mamba, but the selectivity mechanism is not isolated from added parameter capacity; worth a real review. read the letter →

arxiv 2412.08947 v1 pith:2VTWB2FF submitted 2024-12-12 cs.CV cs.AI

Yifeng Yao , Zichen Liu , Zhenyu Cui , Yuxin Peng , Jiahuan Zhou This is my paper · ORCID
classification cs.CVcs.AI
keywords SelectiveVisualPromptingVisionMambastatespacemodelsparameter-efficientfine-tuningprompttuningupdateandforgetgatesHTAbenchmarkVTAB-1K
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

The paper sets out to show that the usual recipe for visual prompting—prepending a few learned tokens to the image sequence—is the wrong interface for Vision Mamba, because Mamba compresses information token by token through input-dependent gates. Its proposed method, Selective Visual Prompting (SVP), instead generates a prompt for every image token at every layer from that token itself, using two lightweight generators: one shared across layers and one per layer. These prompts are added to the tokens before they enter Mamba blocks, so the update and forget gates are computed from prompted inputs and can be steered to keep discriminative information. If the claim is right, a frozen Vim-Small backbone with roughly 1.6M tunable parameters reaches 89.8% average accuracy on HTA and 70.1% on VTAB-1K, beating prefix-prompt and image-prompt baselines. That would make Mamba-style backbones practically adaptable without full fine-tuning.

What carries the argument

The load-bearing object is the selective prompter, a lightweight generator attached to each Mamba block. Cross-Prompting uses one fully connected generator with parameters shared across a group of layers (4, 8, or 12) to produce prompts that carry common information; Inner-Prompting uses a per-layer bottleneck MLP—linear down to 64 dimensions, SiLU activation, linear up—to produce layer-specific prompts. Two zero-initialized elementwise scaling vectors, $\alpha$ and $\beta$, learn how much of each path to add, and the combined prompt is added directly to every image token before the block. This machinery matters because it changes the inputs to the functions $S_B$, $S_C$, and $S_\Delta$ that generate Mamba's data-dependent gates, so the prompt controls, at every token, which information is updated into and forgotten from the recurrent state.

What would settle it

Build a non-selective version of SVP with the same parameter budget and dual-path structure in which each prompt is a learned constant per layer (or per layer group) rather than a function of the input, and compare on HTA and VTAB-1K; if it matches SVP's accuracy, the claim that input-dependent gate activation is the cause is refuted.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that input-dependent, token-wise prompts can activate the selective gates of Vision Mamba in a way that prefixed prompt tokens cannot. For each token $x_i$, SVP forms $\bar{p}_i = \alpha \odot p^C_i + \beta \odot p^I_i$ from a cross-layer prompt $p^C_i$ and an inner-layer prompt $p^I_i$, then feeds $x_i + \bar{p}_i$ into the Mamba block. Because Mamba computes its input-dependent parameters $B_i$, $C_i$, and $\Delta_i$ from the prompted token, the recurrence's update term $B_i(\Delta_i \odot x^p_i)$ and forget factor $\exp(S_\Delta(x^p_i) \odot A)$ are directly affected by the prompt at every sequence position. The paper argues that this selective activation is what lets the model write discriminative features into the hidden state and discard distractions, and it supports the argument with accuracy gains on two benchmarks and with visualizations of update-gate activity across layers.

Load-bearing premise

The paper assumes that the accuracy gains come from input-dependent selectivity of the prompters steering Mamba's gates, rather than from the extra per-layer trainable capacity those prompters add; it reports no parameter-matched non-selective baseline.

Editorial extensions

If this is right

  • Frozen Vision Mamba backbones can be competitively adapted with about 1.6M trainable parameters, outperforming full fine-tuning on 7 of 10 HTA datasets in aggregate and matching much larger ViT-B prompting baselines on several datasets.
  • Prefix-based and image-overlay prompting methods are not the right inductive bias for sequential state-space vision models; token-wise, input-dependent prompts are.
  • Separating shared cross-layer prompts from per-layer inner prompts captures complementary information, and tuning the balance between them with learned scaling vectors improves accuracy over either path alone.
  • The largest VTAB-1K gains appear in the Structured group (+13.8 points over VPT), suggesting selective gate activation especially helps tasks that require spatial and relational reasoning.

Reading between the lines

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

  • Because the gating mechanism SVP exploits is generic to selective state-space models, the same prompt design should transfer to larger or bidirectional Mamba vision backbones beyond the Vim-Small tested here; this is an extrapolation, not a result in the paper.
  • A parameter-matched control with input-independent per-layer prompts would separate the contribution of selectivity from the contribution of added capacity; without that control, the mechanistic story is not fully isolated.
  • Since $\alpha$ and $\beta$ start at zero, SVP preserves the pre-trained model's output at initialization, which suggests it could be composed with other parameter-efficient adapters such as LoRA without disturbing the base model's behavior.
  • The update-gate visualizations could be turned into a quantitative diagnostic: one could measure how much SVP shifts gate statistics per layer and per task, and use that as a predictor of when selective prompting helps.
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

5 major / 6 minor

Summary. The paper proposes Selective Visual Prompting (SVP), a parameter-efficient fine-tuning method tailored to Vision Mamba (Vim). Instead of appending a fixed prompt sequence, SVP adds lightweight per-layer generators that produce token-wise prompts from the input image tokens: a shared Cross-Prompting generator for cross-layer information and per-layer Inner-Prompting generators for layer-specific information, combined through learnable element-wise scaling factors. The prompted tokens are fed into frozen Vim blocks, with only the prompt generators, scaling factors, and classification head trained. The authors evaluate on HTA (10 datasets) and VTAB-1K (19 datasets) with a Vim-Small backbone, reporting average accuracies of 89.8% on HTA and 70.1% on VTAB-1K, surpassing VPT, DAM-VP, SPT, and AutoVP baselines. They also provide ablations of prompt format/position, component contributions, hyperparameters, and visualizations of update gate activations. The paper claims this is the first visual prompting method specifically designed for Vim and attributes the gains to input-dependent selectivity that activates Vim's update and forget gates.

Significance. If the reported results hold, SVP would be a useful and practical PEFT method for Mamba-style vision backbones, and the paper would be one of the first to explicitly address the sequential token-wise processing of Vim in a prompting framework. Strengths include broad benchmark coverage on HTA and VTAB-1K, a publicly released codebase, ablations of the main design choices, and gate-level visualizations that connect the method to the internal mechanism of Mamba. The main weakness is that the central mechanistic claim—that input-dependent selectivity, rather than added per-layer capacity, drives the improvements—is not isolated by any parameter-matched control. Because the proposed generators are essentially per-layer low-rank transformations, the reported gains could in principle be reproduced by a non-selective per-layer adapter or static prompt generator with the same parameter budget. The theoretical discussion in Equations (11)–(12) is substitutional and does not by itself prove the selectivity mechanism.

major comments (5)
  1. [§3, Eqs. (4)–(7); Tables 1–2; Table 4] The central claim that input-dependent selectivity drives the gains is not isolated from added parameter capacity. SVP adds per-layer Inner-Prompting generators (linear down/up with SiLU) plus a shared Cross-Prompting generator, totaling roughly 1.6M tunable parameters, whereas the VPT-deep baseline on Vim-S adds only prompt-token parameters on the order of 9K. The ablations in Table 4 remove entire components, so they vary input-dependence and capacity simultaneously. I request a parameter-matched control: for example, a non-selective variant in which each layer's prompt is a single learned vector, or a per-layer low-rank adapter on x_i with the same rank and parameter budget that is not a function of the input. Without such a control, the improved accuracy in Tables 1 and 2 could be explained by the additional per-layer low-rank transform alone, which would undermine the paper's mechanistic interpretation.
  2. [§4, Eqs. (11)–(12) and Discussion] The theoretical account in Equations (11)–(12) is a substitution: it rewrites the Mamba recurrence after replacing x_i with x_i + \bar p_i. It shows that the prompted input appears in the input-dependent B_i, C_i, and \Delta_i maps, but it does not establish that this promotes discriminative information propagation, nor does it distinguish selectivity from any other input perturbation. Since B_i, C_i, and \Delta_i are input-dependent by construction, the statement that prompts activate the gates is partly tautological and not falsifiable as stated. Please either provide a precise statement—for example, a condition on gate magnitudes or hidden-state norms—or reframe the discussion as an empirical observation supported by controlled experiments rather than a derivation.
  3. [Implementation Details and Figure 4] The number of shared layers in Cross-Prompting is reported as "set to 4, 8, or 12, depending on the dataset," but no selection rule is described. If this hyperparameter is tuned per dataset using the test set, the reported averages may overstate generalization. Please specify whether the choice is made on a validation split and with what criterion, report the chosen value for each of the 29 datasets, and show the sensitivity of the final averages to this choice. The same issue applies to the hidden dimension of the Inner-Prompting generator, which is set to 64 after observing the trend in Figure 5; a selection rule should be stated.
  4. [Tables 1 and 2] All accuracies are reported without error bars or number of seeds. On several datasets the differences are small (for example, SVP vs. DAM-VP on GTSRB is 97.5 vs. 96.5, and on CIFAR10 it is 98.6 vs. 98.0), and single-run results are insufficient to support the claim of significant improvement. Please report mean and standard deviation over at least three independent runs for the main comparisons, or otherwise provide a variance estimate.
  5. [Table 2] The VTAB-1K comparison includes only VPT and SPT among prompting baselines, while the HTA table includes DAM-VP, E2VPT, AutoVP, and SA2VP. Since the abstract and conclusion claim superiority over state-of-the-art visual prompting methods, the VTAB-1K evaluation should either include the same set of baselines or explicitly state why they are omitted. Without this, the breadth of the SOTA claim is not fully supported.
minor comments (6)
  1. [§4, paragraph before Eq. (11)] The word "pracatically" should be "practically".
  2. [Ablation Study, 'Ablation of Prompt Format and Position'] The word "sequencial" should be "sequential".
  3. [Abstract and Introduction] The manuscript alternates between "update and output gates" and "update and forget gates"; please use consistent terminology, preferably matching the Mamba formulation used in the paper.
  4. [Figures 4 and 5] The text inside these figures appears corrupted or poorly rendered (e.g., unicode fragments), and the axes are not labeled clearly. Please provide clean vector figures with readable legend and axis labels.
  5. [Figures 6 and 7] The update-gate visualizations show differences between methods, but without a static-prompt or parameter-matched control they cannot distinguish the selectivity mechanism from added capacity; if these figures are meant to support the mechanism, a quantitative summary (e.g., mean gate values over layers) and the requested control would strengthen the interpretation.
  6. [Overall] The paper does not report the computational overhead of generating token-wise prompts at every layer; adding a sentence on latency or FLOPs relative to the frozen backbone would help practitioners assess the efficiency claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: SVP's reported gains are empirical benchmark results, and Equations 11-12 are an algebraic substitution, not a prediction derived from fitted constants.

full rationale

The paper's only formal derivation is in the 'Discussion and Analysis' section, where Equations 11 and 12 substitute the prompted input x_i + p̄_i into the Mamba recurrence. Because Mamba's B_i, C_i, and Δ_i are defined as input-dependent functions, the rewritten recurrence follows by substitution; it is an identity, not a circular prediction. The mechanism claim that prompts 'activate' the input-dependent gates is true by the architecture's definition, but it is not used to derive the accuracy numbers. The main empirical claims (Tables 1 and 2) compare SVP against external baselines on public HTA and VTAB-1K benchmarks, and the reported accuracies are measured outcomes, not quantities reconstructed from the training objective or from fitted parameters. The ablations in Tables 3 and 4 vary prompt format, position, and components, but the absence of a parameter-matched non-selective control is a potential confound about whether input-dependent selectivity or added capacity drives the gains; it is an experimental-design limitation, not circularity. No load-bearing self-citation or imported uniqueness theorem is present, and no result is renamed from a known pattern. Accordingly, no circular step can be exhibited with the required equation-level reduction, so the appropriate finding is no significant circularity.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

SVP introduces no new physical entities. The central claim rests on two tuned hyperparameters (inner generator width and number of layers sharing the cross generator) plus standard PEFT assumptions about frozen backbones. The dual-path motivation is a paper-specific design assumption rather than a proven property of Vim.

free parameters (2)
  • hidden_dim of Inner-Prompting generator = 64
    Chosen after ablation (Figure 5) as a trade-off between accuracy and parameter count; performance keeps rising with larger values, so the exact value is a tuned capacity hyperparameter.
  • number of shared layers in Cross-Prompting = 4, 8, or 12 depending on dataset
    Implementation Details state the value is dataset-dependent, but the selection rule is not described; this is a per-dataset tuned hyperparameter that affects results.
assumptions (3)
  • domain assumption Frozen pre-trained Vim features are sufficient for downstream tasks when combined with shallow prompts.
    The whole method freezes the backbone and only trains prompt generators and head; this is common in PEFT but is an empirical premise, not proven.
  • ad hoc to paper Vim propagates both shared cross-layer information and specific inner-layer information, and these two types are best handled by separate generators.
    This motivates the dual-path design (Cross-Prompting and Inner-Prompting) but is not formally justified; it is a design assumption specific to this paper.
  • standard math Adding a prompt to x_i changes the input-dependent B_i, C_i, Delta_i and therefore 'activates' the gates.
    Given the Mamba definition (Equations 10-11), any additive input change alters the gate inputs by construction; this is algebraically true but does not by itself imply better discriminative propagation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Selective Visual Prompting in Vision Mamba." pith.science (2026). https://pith.science/paper/2VTWB2FF

@misc{pith2026241208947,
  author       = {Pith},
  title        = {Pith review of: Selective Visual Prompting in Vision Mamba},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2VTWB2FF}},
  note         = {Machine review of arXiv:2412.08947}
}
read the original abstract

Pre-trained Vision Mamba (Vim) models have demonstrated exceptional performance across various computer vision tasks in a computationally efficient manner, attributed to their unique design of selective state space models. To further extend their applicability to diverse downstream vision tasks, Vim models can be adapted using the efficient fine-tuning technique known as visual prompting. However, existing visual prompting methods are predominantly tailored for Vision Transformer (ViT)-based models that leverage global attention, neglecting the distinctive sequential token-wise compression and propagation characteristics of Vim. Specifically, existing prompt tokens prefixed to the sequence are insufficient to effectively activate the input and forget gates across the entire sequence, hindering the extraction and propagation of discriminative information. To address this limitation, we introduce a novel Selective Visual Prompting (SVP) method specifically for the efficient fine-tuning of Vim. To prevent the loss of discriminative information during state space propagation, SVP employs lightweight selective prompters for token-wise prompt generation, ensuring adaptive activation of the update and forget gates within Mamba blocks to promote discriminative information propagation. Moreover, considering that Vim propagates both shared cross-layer information and specific inner-layer information, we further refine SVP with a dual-path structure: Cross-Prompting and Inner-Prompting. Cross-Prompting utilizes shared parameters across layers, while Inner-Prompting employs distinct parameters, promoting the propagation of both shared and specific information, respectively. Extensive experimental results on various large-scale benchmarks demonstrate that our proposed SVP significantly outperforms state-of-the-art methods. Our code is available at https://github.com/zhoujiahuan1991/AAAI2025-SVP.

Figures

Figures reproduced from arXiv: 2412.08947 by the authors.

Figure 1
Figure 1. Existing visual prompting methods (Jia et al. 2022) [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Our SVP employs a dual-path architecture: the Inner-Prompting pathway prompts specific information at each layer, [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. The internal structure of the Mamba block. The [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Ablation results of the number of shared layers in [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Ablation of hidden dimension in Inner-Prompting. [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: Visualization of normalized update gate of Vim [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 13 canonical work pages

  1. [1]

    , " * 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 mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    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 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Basu, S.; Hu, S.; Massiceti, D.; and Feizi, S. 2024. Strong Baselines for Parameter-Efficient Few-Shot Fine-Tuning. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 11024--11031

    work page 2024

  4. [4]

    Bossard, L.; Guillaumin, M.; and Van Gool, L. 2014. Food-101--mining discriminative components with random forests. In Computer vision--ECCV 2014: 13th European conference, zurich, Switzerland, September 6-12, 2014, proceedings, part VI 13, 446--461. Springer

    2014

  5. [5]

    Chen, S.; Ge, C.; Tong, Z.; Wang, J.; Song, Y.; Wang, J.; and Luo, P. 2022. Adaptformer: Adapting vision transformers for scalable visual recognition. Advances in Neural Information Processing Systems, 35: 16664--16678

    2022

  6. [6]

    Cimpoi, M.; Maji, S.; Kokkinos, I.; Mohamed, S.; and Vedaldi, A. 2014. Describing textures in the wild. In Proceedings of the IEEE conference on computer vision and pattern recognition, 3606--3613

    2014

  7. [7]

    Dao, T.; and Gu, A. 2024. Transformers are SSM s: Generalized Models and Efficient Algorithms Through Structured State Space Duality. In Forty-first International Conference on Machine Learning

    2024

  8. [8]

    Deng, J.; Dong, W.; Socher, R.; Li, L.-J.; Li, K.; and Fei-Fei, L. 2009. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition, 248--255. Ieee

    2009

Show all 41 references
  1. [9]

    Dong, W.; Yan, D.; Lin, Z.; and Wang, P. 2024. Efficient adaptation of large vision transformer via adapter re-composing. Advances in Neural Information Processing Systems, 36

    2024

  2. [10]

    Dosovitskiy, A.; Beyer, L.; Kolesnikov, A.; Weissenborn, D.; Zhai, X.; Unterthiner, T.; Dehghani, M.; Minderer, M.; Heigold, G.; Gelly, S.; et al. 2020. An image is worth 16x16 words: Transformers for image recognition at scale. arXiv preprint arXiv:2010.11929

    2020 arXiv

  3. [11]

    Elfwing, S.; Uchibe, E.; and Doya, K. 2018. Sigmoid-weighted linear units for neural network function approximation in reinforcement learning. Neural networks, 107: 3--11

    2018

  4. [12]

    M.-C.; Lam, W.; Bing, L.; and Collier, N

    Fu, Z.; Yang, H.; So, A. M.-C.; Lam, W.; Bing, L.; and Collier, N. 2023. On the effectiveness of parameter-efficient fine-tuning. In Proceedings of the AAAI conference on artificial intelligence, volume 37, 12799--12807

    2023

  5. [13]

    Gu, A.; and Dao, T. 2023. Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752

    2023 arXiv

  6. [14]

    Gu, A.; Goel, K.; and R \'e , C. 2021. Efficiently modeling long sequences with structured state spaces. arXiv preprint arXiv:2111.00396

    2021 arXiv

  7. [15]

    Han, C.; Wang, Q.; Cui, Y.; Cao, Z.; Wang, W.; Qi, S.; and Liu, D. 2023. E\^ 2VPT: An Effective and Efficient Approach for Visual Prompt Tuning. In Proceedings of the IEEE/CVF International Conference on Computer Vision, 17491--17502

    2023

  8. [16]

    Huang, Q.; Dong, X.; Chen, D.; Zhang, W.; Wang, F.; Hua, G.; and Yu, N. 2023. Diversity-aware meta visual prompting. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 10878--10887

    2023

  9. [17]

    Jia, M.; Tang, L.; Chen, B.-C.; Cardie, C.; Belongie, S.; Hariharan, B.; and Lim, S.-N. 2022. Visual prompt tuning. In European Conference on Computer Vision, 709--727. Springer

    2022

  10. [18]

    Khosla, A.; Jayadevaprakash, N.; Yao, B.; and Li, F.-F. 2011. Novel dataset for fine-grained image categorization: Stanford dogs. In Proc. CVPR workshop on fine-grained visual categorization (FGVC), volume 2

    2011

  11. [19]

    Kornblith, S.; Shlens, J.; and Le, Q. V. 2019. Do better imagenet models transfer better? In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2661--2671

    2019

  12. [20]

    Krizhevsky, A.; Hinton, G.; et al. 2009. Learning multiple layers of features from tiny images

    2009

  13. [21]

    Krizhevsky, A.; Sutskever, I.; and Hinton, G. E. 2012. Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems, 25

    2012

  14. [22]

    Liu, Z.; Peng, Y.; and Zhou, J. 2024. Ins VP : Efficient Instance Visual Prompting from Image Itself. In ACM Multimedia 2024

    2024

  15. [23]

    Loshchilov, I.; and Hutter, F. 2017. Decoupled weight decay regularization. arXiv preprint arXiv:1711.05101

    2017 arXiv

  16. [24]

    Y.; et al

    Netzer, Y.; Wang, T.; Coates, A.; Bissacco, A.; Wu, B.; Ng, A. Y.; et al. 2011. Reading digits in natural images with unsupervised feature learning. In NIPS workshop on deep learning and unsupervised feature learning, volume 2011, 4. Granada

    2011

  17. [25]

    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

  18. [26]

    Pechlivanidou, G.; and Karampetakis, N. 2022. Zero-order hold discretization of general state space systems with input delay. IMA Journal of Mathematical Control and Information, 39(2): 708--730

    2022

  19. [27]

    Pei, W.; Xia, T.; Chen, F.; Li, J.; Tian, J.; and Lu, G. 2024. SA ^2 VP: Spatially Aligned-and-Adapted Visual Prompt. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 4450--4458

    2024

  20. [28]

    Russakovsky, O.; Deng, J.; Su, H.; Krause, J.; Satheesh, S.; Ma, S.; Huang, Z.; Karpathy, A.; Khosla, A.; Bernstein, M.; et al. 2015. Imagenet large scale visual recognition challenge. International journal of computer vision, 115: 211--252

    2015

  21. [29]

    Stallkamp, J.; Schlipsing, M.; Salmen, J.; and Igel, C. 2012. Man vs. computer: Benchmarking machine learning algorithms for traffic sign recognition. Neural networks, 32: 323--332

    2012

  22. [30]

    O.; and Roth, S

    Steitz, J.-M. O.; and Roth, S. 2024. Adapters Strike Back. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 23449--23459

    2024

  23. [31]

    Sung, Y.-L.; Cho, J.; and Bansal, M. 2022. Lst: Ladder side-tuning for parameter and memory efficient transfer learning. Advances in Neural Information Processing Systems, 35: 12991--13005

    2022

  24. [32]

    Touvron, H.; Cord, M.; El-Nouby, A.; Verbeek, J.; and J \'e gou, H. 2022. Three things everyone should know about vision transformers. In European Conference on Computer Vision, 497--515. Springer

    2022

  25. [33]

    Tsao, H.-A.; Hsiung, L.; Chen, P.-Y.; Liu, S.; and Ho, T.-Y. 2024. Auto VP : An Automated Visual Prompting Framework and Benchmark. In The Twelfth International Conference on Learning Representations

    2024

  26. [34]

    Van Horn, G.; Branson, S.; Farrell, R.; Haber, S.; Barry, J.; Ipeirotis, P.; Perona, P.; and Belongie, S. 2015. Building a bird recognition app and large scale dataset with citizen scientists: The fine print in fine-grained dataset collection. In Proceedings of the IEEE confer...

    2015

  27. [35]

    Wah, C.; Branson, S.; Welinder, P.; Perona, P.; and Belongie, S. 2011. The caltech-ucsd birds-200-2011 dataset

    2011

  28. [36]

    Wang, Y.; Cheng, L.; Fang, C.; Zhang, D.; Duan, M.; and Wang, M. 2024. Revisiting the Power of Prompt for Visual Tuning. In Forty-first International Conference on Machine Learning

    2024

  29. [37]

    Xin, Y.; Du, J.; Wang, Q.; Lin, Z.; and Yan, K. 2024 a . Vmt-adapter: Parameter-efficient transfer learning for multi-task dense scene understanding. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 16085--16093

    2024

  30. [38]

    Xin, Y.; Luo, S.; Zhou, H.; Du, J.; Liu, X.; Fan, Y.; Li, Q.; and Du, Y. 2024 b . Parameter-efficient fine-tuning for pre-trained vision models: A survey. arXiv preprint arXiv:2402.02242

    2024

  31. [39]

    B.; Ravfogel, S.; and Goldberg, Y

    Zaken, E. B.; Ravfogel, S.; and Goldberg, Y. 2021. Bitfit: Simple parameter-efficient fine-tuning for transformer-based masked language-models. arXiv preprint arXiv:2106.10199

    2021

  32. [40]

    S.; Neumann, M.; Dosovitskiy, A.; et al

    Zhai, X.; Puigcerver, J.; Kolesnikov, A.; Ruyssen, P.; Riquelme, C.; Lucic, M.; Djolonga, J.; Pinto, A. S.; Neumann, M.; Dosovitskiy, A.; et al. 2019. A large-scale study of representation learning with the visual task adaptation benchmark. arXiv preprint arXiv:1910.04867

    2019 arXiv

  33. [41]

    Zhu, L.; Liao, B.; Zhang, Q.; Wang, X.; Liu, W.; and Wang, X. 2024. Vision Mamba: Efficient Visual Representation Learning with Bidirectional State Space Model. In Forty-first International Conference on Machine Learning

    2024

Pith tools

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