Pith. sign in

REVIEW 3 major objections 6 minor 53 references

Full Conformal Adaptation of Medical Vision-Language Models

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read Full conformal adaptation gives medical vision-language models the same coverage guarantees as split conformal prediction while producing prediction sets up to 27% smaller, using exactly the same few-shot data.

desk verdict A genuinely useful transductive conformal adaptation framework for few-shot medical VLMs, undercut mostly by experiments that never quite test the coverage guarantee in the regime where it is supposed to hold. read the letter →

arxiv 2506.06076 v1 pith:S6DPHEMN submitted 2025-06-06 cs.CV

Julio Silva-Rodríguez , Leo Fillioux , Paul-Henry Cournède , Maria Vakalopoulou , Stergios Christodoulidis , Ismail Ben Ayed , Jose Dolz This is my paper
classification cs.CV
keywords conformalpredictionvision-languagemodelsfew-shotadaptationtransductivelearningmedicalimageanalysisuncertaintyquantificationlinearprobingcoverageguarantee
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's central claim is that a pre-trained medical vision-language model can be adapted to a new task and conformalized in a single transductive procedure—full conformal adaptation (FCA)—using exactly the same few-shot labeled support set that split conformal prediction would spend only on calibration. If this works as stated, reliability guarantees and few-shot transfer no longer compete for scarce labeled medical data: the model both learns the new task and produces prediction sets that cover the true label with the demanded frequency. The authors report that FCA matches the accuracy of standard adaptation, preserves marginal coverage at the nominal level, and produces sets up to 27% smaller than split conformal prediction across nine medical classification tasks spanning histology, retina, and chest X-ray. They also introduce SS-Text, a closed-form linear probe solver that sidesteps gradient training, making the per-test-point adaptation computationally feasible.

What carries the argument

The central object is full conformal adaptation (FCA), a transductive conformal procedure in which each test image, paired in turn with each candidate label, is added to a few-shot adaptation set while a linear probe is fit over frozen VLM features; the per-label quantile of non-conformity scores then defines the prediction set. Its validity is carried by the exchangeability of the adaptation set with each test point, together with permutation invariance of the fitting procedure, plus the standard full-conformal argument. The supporting machinery is SS-Text, a closed-form solution to a constrained linear probe objective whose minimizer is a weighted sum of the class text prototype and the mean of the per-class visual features, with the regularization weight set to $1/(N\tau)$ so that the textual prior fades as support data grows.

What would settle it

Split a dataset by patient so no patient appears in both the support/adaptation set and the test set, then run FCA with K=16 and α=0.10; if observed coverage falls below roughly 0.90 by more than sampling error, the coverage claim is violated. The paper's own SICAPv2 and MMAC results already show such gaps, so the test would quantify how far the guarantee degrades under patient-level shift.

Watch

Extended reading notes

Core claim

In classical full conformal prediction, the model is retrained for each test point and each candidate label, using that test point's hypothetical label alongside training data, so that scores for the true label are exchangeable with training scores. The paper transplants this logic into few-shot VLM adaptation: instead of retraining a large network, it fits a lightweight linear probe on top of frozen VLM features, once for each candidate label of each test image, with the adaptation set replacing the pre-training dataset as the exchangeable reference. The result is a prediction set per test point that, the paper argues, inherits FCP's finite-sample marginal coverage guarantee while using exactly the same data as SCP. Empirically, across three medical VLMs and nine tasks, FCA maintains coverage at the target 1-α while cutting average set size relative to SCP by up to 27%, and it improves per-class coverage gaps substantially. A training-free solver, SS-Text, solves the constrained linear-probe objective in closed form and is about 150× faster per fit than gradient-based adaptation, making the transductive loop practical.

Load-bearing premise

The guarantee holds only if the few-shot adaptation set is exchangeable with each new test image and the probe fitting is permutation-invariant; when test patients differ systematically from the support set, the marginal coverage promise can fail.

Editorial extensions

If this is right

  • FCA lets a single few-shot support set do double duty—adaptation and conformal calibration—so medical deployments no longer need separate calibration subsets to get coverage guarantees.
  • At the claimed 27% set-size reduction, clinicians using FCA would on average read fewer candidate labels per image while retaining the same 90% or 95% coverage promise.
  • The closed-form SS-Text solver removes the computational objection to transductive conformal prediction, since the per-label fits cost milliseconds rather than gradient-training runs.
  • Across 9 tasks and 3 modality-specialized VLMs, the framework gives better class-conditional coverage than Adapt+SCP, so underrepresented classes are less likely to be dropped from the sets.
  • The coverage guarantee holds at the marginal level and is not exact per patient or per class; patient-shifted datasets such as SICAPv2 and MMAC show coverage gaps even for the proposed method.

Reading between the lines

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

  • The same transductive adaptation trick could apply to other foundation-model families with frozen features and a cheap probe, opening a general route to conformalized few-shot transfer beyond VLMs.
  • Because the validity argument lives in feature space, coverage might be maintained even when the raw image distribution shifts, as long as the adapted features of new patients remain interchangeable with the support features—an empirically testable condition.
  • The closed-form solver's $\lambda = 1/(N\tau)$ schedule could be replaced by a data-driven per-class weight, potentially improving low-shot accuracy and coverage on grading tasks where inter-observer variability is high.
  • For dense prediction, the per-label per-image fits remain a bottleneck; making the probe fitting incremental or amortized would be the natural next step to extend FCA beyond image-level classification.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper proposes full conformal adaptation (FCA), a transductive conformal prediction framework for adapting medical vision-language models (VLMs) using a few-shot adaptation set. For each test point, FCA fits a linear probe on the adaptation set augmented with each candidate label, computes label-wise non-conformity scores, and forms a prediction set via the full conformal quantile. The authors also introduce SS-Text, a training-free solver that approximates a constrained linear probing objective by minimizing only the hard-label term. Experiments on 9 public datasets and 3 medical VLMs report that FCA maintains coverage near nominal while reducing set sizes by up to 27% relative to split conformal prediction, and that SS-Text is about 150× faster than gradient-based adapters.

Significance. If validated, FCA would be a practical way to combine few-shot VLM adaptation with distribution-free finite-sample coverage guarantees while using the same labeled data as SCP, and SS-Text addresses a real computational bottleneck of transductive conformal prediction. The paper provides code, evaluates across multiple modalities and datasets, and correctly inherits the FCP coverage argument. However, the current experimental protocol does not actually test the coverage guarantee under the exchangeability assumption on which it rests, and key tables lack uncertainty estimates; these issues need to be addressed before the central claim can be accepted.

major comments (3)
  1. [§4.2, Eq. (5); §5.1–5.2; Table 1] The coverage guarantee in Eq. (5) requires the adaptation set D_adapt and each test point to be exchangeable, and the fitting procedure to be permutation-invariant. The experimental protocol in §5.1 uses a class-balanced few-shot support set (N = K·C, as stated in §4.1) while the medical test sets are naturally imbalanced; §5.2 concedes that 'the label-marginal distribution of the testing data is unknown, and typical few-shot adaptation pipelines assume a balanced support set.' This violates exchangeability even for random patient splits, so Table 1's coverage numbers (e.g., FCA LAC at α=0.10 reaches 0.896, not 0.900) are obtained in a regime where the guarantee is not supposed to hold. The paper therefore never empirically demonstrates the central claim that FCA maintains nominal coverage under exchangeable data. Please add an experiment with exchangeable test sampling (e.g., balanced test labels or sampling test points from the same class-balanced distribution), or apply a label-shift-robust conformal quantile, and report the resulting coverage.
  2. [Table 1] Despite stating that all results are repeated over 20 random seeds, Table 1 reports only point estimates for coverage, set size, and CCV. The differences relative to nominal coverage are small (e.g., 0.896 vs 0.900 for FCA-LAC at α=0.10), and without standard deviations or confidence intervals it is impossible to tell whether the deficit is statistically significant or an artifact of finite sampling. Please report means with error bars (or confidence intervals) for all metrics in Tables 1 and 2 and for the per-dataset results in Fig. 2.
  3. [§4.3, Eqs. (6)–(10)] The derivation of SS-Text minimizes only the g1 term of the loss in Eq. (6) and sets aside g2, the log-sum-exp term that couples classes through the softmax denominator. The text calls the resulting closed-form solution an approximation, but no bound on the approximation error is provided, and the experiments do not quantify how far the solution is from the true minimizer of Eq. (6). Since SS-Text is a central contribution and is used for all FCA experiments, please either justify the neglect of g2 (e.g., show that under λ = 1/(Nτ) the g2 gradient is small or the objective is dominated by g1) or explicitly present SS-Text as a heuristic and discuss the sensitivity of the downstream conformal results to this approximation.
minor comments (6)
  1. [Abstract and §5.2] The phrase 'same coverage guarantees' is stronger than what the experiments show; consider saying 'target coverage' or adding the caveat that under exchangeability the guarantee holds, while in the evaluated label-shifted setting coverage is approximate.
  2. [Fig. 1 caption] The statement that FCA 'translates the exchangeability assumption to the feature space' is imprecise; features are deterministic transforms of images, so exchangeability is not translated but inherited, and the converse does not follow.
  3. [Eq. (10)] Please specify whether the resulting class vectors w*_c are ℓ2-normalized before being used in Eq. (1); the text normalizes v and t, but the sum in Eq. (10) is not automatically normalized.
  4. [References] References [22] and [23] are duplicate entries of the same Lei et al. paper; one should be removed.
  5. [Table 2] The header 'SS-Text λc ≃ zero-shot perf.' is cryptic; it refers to CLAP's class-wise multiplier from [39], but the notation is unexplained in the text.
  6. [Discussion] The phrase 'tenths of tags' should be 'tens of tags'.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: FCA's coverage guarantee is imported from established FCP results and the empirical comparisons are external.

full rationale

FCA's validity step (Section 4.2, Eq. 5) is the standard full conformal prediction quantile rule, and the paper explicitly cites the original FCP works [1,36,31] rather than deriving a new guarantee; this is legitimate external support, not a self-citation chain. SS-Text (Eq. 10) is an analytic minimizer of a surrogate of Eq. 6 with lambda fixed to 1/(N tau) by design, and Table 2 compares this choice against other lambda values; it is not fitted to the coverage or set-size metrics reported in Table 1. The empirical benchmarks use public datasets (CONCH, FLAIR, and CONVIRT features; NCT-CRC, SICAPv2, MESSIDOR, CheXpert, and others) and standard baselines, so the efficiency and coverage comparisons are not forced by construction. The self-citations to CLAP, LP++, and FLAIR are baseline implementations and model choices, not load-bearing premises of the conformal argument. The paper itself acknowledges in Section 5.2 and the Discussion that balanced few-shot support sets and patient-level shifts can violate exchangeability; this is an honest limitation affecting external validity in some datasets, but it is not a circular derivation. No equation in the paper reduces a predicted quantity to the same fitted quantity by definition.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The central coverage claim rests on the standard FCP theorem (exchangeability plus permutation invariance) and on the empirical adequacy of SS-Text as an approximation. No new physical or conceptual entities are introduced. The only hand-set parameter is lambda, which controls the text/visual trade-off in SS-Text.

free parameters (1)
  • lambda_SS-Text = 1/(N*tau)
    Hand-chosen schedule in Eq. 10 for SS-Text, balancing text prototype and visual features. It affects set efficiency but not coverage validity. Table 2 compares fixed values and supports the chosen schedule.
assumptions (4)
  • domain assumption Exchangeability of adaptation set and test points
    Required for the FCP-style coverage guarantee invoked in Section 4.1. Acknowledged as potentially unrealistic in medical settings (Section 5.2 and Discussion).
  • standard math Permutation invariance of the fitting procedure
    The FCP guarantee holds only if the model fit is symmetric in the augmented data. SS-Text is symmetric because it uses sums, but the paper does not prove this property explicitly.
  • domain assumption Linear probe sufficiency on frozen VLM features
    The method adapts only a linear classifier over pre-trained features (Section 4.2), assuming the features are transferable enough. This is common in the few-shot VLM literature.
  • ad hoc to paper SS-Text approximates the constrained linear probing objective
    Eq. 10 minimizes only g1 and ignores g2, the softmax coupling term. The adequacy of this approximation is shown empirically (Table 2), not theoretically.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Full Conformal Adaptation of Medical Vision-Language Models." pith.science (2026). https://pith.science/paper/S6DPHEMN

@misc{pith2026250606076,
  author       = {Pith},
  title        = {Pith review of: Full Conformal Adaptation of Medical Vision-Language Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S6DPHEMN}},
  note         = {Machine review of arXiv:2506.06076}
}
read the original abstract

Vision-language models (VLMs) pre-trained at large scale have shown unprecedented transferability capabilities and are being progressively integrated into medical image analysis. Although its discriminative potential has been widely explored, its reliability aspect remains overlooked. This work investigates their behavior under the increasingly popular split conformal prediction (SCP) framework, which theoretically guarantees a given error level on output sets by leveraging a labeled calibration set. However, the zero-shot performance of VLMs is inherently limited, and common practice involves few-shot transfer learning pipelines, which cannot absorb the rigid exchangeability assumptions of SCP. To alleviate this issue, we propose full conformal adaptation, a novel setting for jointly adapting and conformalizing pre-trained foundation models, which operates transductively over each test data point using a few-shot adaptation set. Moreover, we complement this framework with SS-Text, a novel training-free linear probe solver for VLMs that alleviates the computational cost of such a transductive approach. We provide comprehensive experiments using 3 different modality-specialized medical VLMs and 9 adaptation tasks. Our framework requires exactly the same data as SCP, and provides consistent relative improvements of up to 27% on set efficiency while maintaining the same coverage guarantees.

Figures

Figures reproduced from arXiv: 2506.06076 by the authors.

Figure 1
Figure 1. Overview of conformal frameworks: (a) split conformal prediction [30], (b) Full conformal prediction [1], and (c) full conformal adaptation (Ours). The proposed setting satisfies the coverage guarantees as FCP (see [1,36,31]) but translates the exchangeability assumption to the feature space. Also, note FCA requires exactly the same data sources as SCP, yet is more data efficient, as it allows both adaptation and co… view at source ↗
Figure 2
Figure 2. Conformal prediction results per dataset. Results were obtained us￾ing α = 0.10, and LAC [28]. Each dot represents the performance, with increasing size correlated with the number of shots for adaptation, i.e., K ∈ {4, 8, 16} [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Few-shot adaptation performance (a) and efficiency analysis (b). Efficiency, regarding full conformal adaptation predictions: images per class - per second. The dot size in (b) indicates the number of shots, i.e., K ∈ {4, 8, 16}. 5.4 In-depth studies Role of λ. We fixed λ = 1/(Nτ ) in Eq. 10, assuming that the more supervision signals we leverage for adaptation, the less effect of the constraint is desired [PITH_FU… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Qualitative evaluation of conformal prediction for grading tasks, i.e., Gleason for histology (a) and diabetic retinopathy for retina (b). Left: set size distribution per class. Right: label frequency in sets corresponding for each category. Results using FCA with K = …

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

53 extracted references · 51 canonical work pages

  1. [1]

    In: Conference on Uncertainty in Artificial Intelligence

    Alex Gammerman, Volodya Vovk, V.V.: Learning by transduction. In: Conference on Uncertainty in Artificial Intelligence. pp. 148–156 (1998) 2, 5, 6

    work page 1998

  2. [2]

    In: Clinical Natural Language Processing Workshop (2019) 7

    Alsentzer, E., et al.: Publicly available clinical BERT embeddings. In: Clinical Natural Language Processing Workshop (2019) 7

    work page 2019

  3. [3]

    In: International Conference on Learning Representations (ICLR) (2020) 2, 3, 4, 7, 8

    Angelopoulos, A.N., Bates, S., Jordan, M., Malik, J.: Uncertainty sets for image classifiers using conformal prediction. In: International Conference on Learning Representations (ICLR) (2020) 2, 3, 4, 7, 8

    work page 2020

  4. [4]

    Scientific Reports8 (2018) 12

    Arvaniti, E., Fricker, K., Moret, M., Rupp, N., Hermanns, T., Fankhauser, C., Wey, N., Wild, P., Rüschoff, J., Claassen, M.: Automated gleason grading of prostate cancer tissue microarrays via deep learning. Scientific Reports8 (2018) 12

    work page 2018

  5. [5]

    Chowdhury, M.E.H., Rahman, T., Khandakar, A., Mazhar, R., Kadir, M.A., Mah- bub, Z.B., Islam, K.R., Khan, M.S., Iqbal, A., Emadi, N.A., Reaz, M.B.I., Islam, M.T.: Can ai help in screening viral and covid-19 pneumonia? IEEE Access8, 132665–132676 (2020) 7

    work page 2020

  6. [6]

    In: Advances in Neural Information Process- ing Systems (NeurIPS) (2024) 3, 8

    Correia, A.H., Massoli, F.V., Louizos, C., Behboodi, A.: An information theoretic perspective on conformal prediction. In: Advances in Neural Information Process- ing Systems (NeurIPS) (2024) 3, 8

    work page 2024

  7. [7]

    Image Analysis & Stereology33, 231–234 (07 2014) 7

    Decencière, E., et al.: Feedback on a publicly distributed image database: The messidor database. Image Analysis & Stereology33, 231–234 (07 2014) 7

    work page 2014

  8. [8]

    In: Advances in Neural In- formation Processing Systems (NeurIPS)

    Ding, T., Angelopoulos, A., Bates, S., Jordan, M., Tibshirani, R.J.: Class- conditional conformal prediction with many classes. In: Advances in Neural In- formation Processing Systems (NeurIPS). vol. 36 (2023) 4, 8

    work page 2023

Show all 53 references
  1. [9]

    In: Advances in Neural Information Pro- cessing Systems (NeurIPS) (2022) 3

    Einbinder, B.S., Romano, Y., Sesia, M., Zhou, Y.: Training uncertainty-aware clas- sifiers with conformalized deep learning. In: Advances in Neural Information Pro- cessing Systems (NeurIPS) (2022) 3

    2022

  2. [10]

    In: Medical Image Computing and Computer Assisted Intervention (MICCAI)

    Galdran, A., Dolz, J., Chakor, H., Lombaert, H., Ayed, I.B.: Cost-sensitive regular- ization for diabetic retinopathy grading from eye fundus images. In: Medical Image Computing and Computer Assisted Intervention (MICCAI). pp. 1–7 (10 2020) 12

    2020

  3. [11]

    International Journal of Computer Vision (2023) 3, 6

    Gao, P., Geng, S., Zhang, R., Ma, T., Fang, R., Zhang, Y., Li, H., Qiao, Y.: Clip-adapter: Better vision-language models with feature adapters. International Journal of Computer Vision (2023) 3, 6

    2023

  4. [12]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023) 3

    Hantao Yao, Rui Zhang, C.X.: Visual-language prompt tuning with knowledge- guided context optimization. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023) 3

    2023

  5. [13]

    In: MICCAI Workshop on Data Augmentation, Labelling, and Imper- fections

    Holste, G., Wang, S., Jiang, Z., Shen, T.C., Shih, G., Summers, R.M., Peng, Y., Wang, Z.: Long-tailed classification of thorax diseases on chest x-ray: A new bench- mark study. In: MICCAI Workshop on Data Augmentation, Labelling, and Imper- fections. pp. 22–32 (2022) 7 14 J. S...

    2022

  6. [14]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Huang, Y., Shakeri, F., Dolz, J., Boudiaf, M., Bahig, H., Ayed, I.B.: Lp++: A surprisingly strong linear probe for few-shot clip. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp. 23773– 23782 (2024) 1, 3, 6, 8, 10, 11, 12

    2024

  7. [15]

    In: Proceedings of the AAAI conference on artificial intelligence

    Irvin, J., Rajpurkar, P., Ko, M., Yu, Y., Ciurea-Ilcus, S., Chute, C., Marklund, H., Haghgoo, B., Ball, R., Shpanskaya, K., et al.: Chexpert: A large chest radiograph dataset with uncertainty labels and expert comparison. In: Proceedings of the AAAI conference on artificial in...

    2019

  8. [16]

    In: International Conference on Machine Learning (ICML)

    Jia, C.,Yang,Y.,Xia,Y.,Chen, Y.T.,Parekh,Z.,Pham, H.,Le,Q.,Sung,Y.H.,Li, Z., Duerig, T.: Scaling up visual and vision-language representation learning with noisy text supervision. In: International Conference on Machine Learning (ICML). pp. 4904–4916 (2021) 1

    2021

  9. [17]

    Scientific Data9, 475 (08 2022) 7

    Jin, K., Huang, X., Zhou, J., Li, Y., Yan, Y., Sun, Y., Zhang, Q., Wang, Y., Ye, J.: Fives: A fundus image dataset for artificial intelligence based vessel segmentation. Scientific Data9, 475 (08 2022) 7

    2022

  10. [18]

    Scientific Data6, 317 (2019) 7

    Johnson, A., et al.: MIMIC-CXR, a de-identified publicly available database of chest radiographs with free-text reports. Scientific Data6, 317 (2019) 7

    2019

  11. [19]

    Zenodo105281 (2018) 7

    Kather, J.N., Halama, N., Marx, A.: 100,000 histological images of human colorec- tal cancer and healthy tissue. Zenodo105281 (2018) 7

    2018

  12. [20]

    Advances in Neural Information Processing Systems 33, 4138– 4149 (2020) 2

    Kim, B., Xu, C., Barber, R.: Predictive inference is free with the jackknife+- after-bootstrap. Advances in Neural Information Processing Systems 33, 4138– 4149 (2020) 2

    2020

  13. [21]

    Frontiers in Oncology12, 1022967 (2022) 7

    Kriegsmann, K., Lobers, F., Zgorzelski, C., Kriegsmann, J., Janssen, C., Meliss, R.R., Muley, T., Sack, U., Steinbuss, G., Kriegsmann, M.: Deep learning for the detection of anatomical tissue structures and neoplasms of the skin on scanned histopathological tissue sections. Fr...

    2022

  14. [22]

    Journal of the American Statistical Association 113(523), 1094–1111 (2018) 4

    Lei, J., G’Sell, M., Rinaldo, A., Tibshirani, R.J., Wasserman, L.: Distribution-free predictive inference for regression. Journal of the American Statistical Association 113(523), 1094–1111 (2018) 4

    2018

  15. [23]

    Journal of the American Statistical Association 113(523), 1094–1111 (2018) 2

    Lei, J., G’Sell, M., Rinaldo, A., Tibshirani, R.J., Wasserman, L.: Distribution-free predictive inference for regression. Journal of the American Statistical Association 113(523), 1094–1111 (2018) 2

    2018

  16. [24]

    In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023) 3, 6

    Lin, Z., Yu, S., Kuang, Z., Pathak, D., Ramanan, D.: Multimodality helps uni- modality: Cross-modal few-shot learning with multimodal models. In: Proceed- ings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2023) 3, 6

    2023

  17. [25]

    Nature Medicine30, 863–874 (2024) 1, 7

    Lu, M.Y., et al.: A visual-language foundation model for computational pathology. Nature Medicine30, 863–874 (2024) 1, 7

    2024

  18. [26]

    In: International Conference on Machine Learning (ICML)

    Luo, H., Bao, J., Wu, Y., He, X., Li, T.: Segclip: Patch aggregation with learnable centers for open-vocabulary semantic segmentation. In: International Conference on Machine Learning (ICML). pp. 23033–23044. PMLR (2023) 1

    2023

  19. [27]

    Nature Methods21 (2024) 8

    Maier-Hein, L., Reinke, A., Godau, P., Tizabi, M.D., Buettner, F., Christodoulou, E., Glocker, B., Isensee, F., Kleesiek, J., Kozubek, M., Reyes, M., Riegler, M., Wiesenfarth, M., Kavur, A., Sudre, C., Baumgartner, M., Eisenmann, M., Heckmann-Nötzel, D., Rädsch, T., Jaeger, P....

    2024

  20. [28]

    Journal of the American Statistical Association114(525), 223–234 (2019) 2, 3, 7, 9 Full Conformal Adaptation of Medical Vision-Language Models 15

    Mauricio Sadinle, J.L., Wasserman, L.: Least ambiguous set-valued classifiers with bounded error levels. Journal of the American Statistical Association114(525), 223–234 (2019) 2, 3, 7, 9 Full Conformal Adaptation of Medical Vision-Language Models 15

    2019

  21. [29]

    In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2023) 3

    Ouali, Y., Bulat, A., Martinez, B., Tzimiropoulos, G.: Black box few-shot adapta- tion for vision-language models. In: Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) (2023) 3

    2023

  22. [30]

    In: European Conference on Machine Learning (ECML)

    Papadopoulos, H., Proedrou, K., Vovk, V., Gammerman, A.: Inductive confidence machines for regression. In: European Conference on Machine Learning (ECML). pp. 345–356 (2002) 2, 4, 6

    2002

  23. [31]

    In: European Conference on Machine Learning (ECML)

    Proedrou, K., Nouretdinov, I., Vovk, V., Gammerman, A.: Transductive confidence machines for pattern recognition. In: European Conference on Machine Learning (ECML). pp. 381–390 (2002) 5, 6

    2002

  24. [32]

    JAMA ophthalmology142(11), 1006–1015 (2024) 7

    Qian, B., et al.: A competition for the diagnosis of myopic maculopathy by artificial intelligence algorithms. JAMA ophthalmology142(11), 1006–1015 (2024) 7

    2024

  25. [33]

    In: International Conference on Machine Learning (ICML)

    Radford, A., Kim, J.W., Hallacy, C., Ramesh, A., Goh, G., Agarwal, S., Sastry, G., Askell, A., Mishkin, P., Clark, J., et al.: Learning transferable visual models from natural language supervision. In: International Conference on Machine Learning (ICML). pp. 8748–8763 (2021) 1...

    2021

  26. [34]

    Computers in Biology and Medicine132, 104319 (2021) 7

    Rahman, T., Khandakar, A., Qiblawey, Y., Tahir, A., Kiranyaz, S., Abul Kashem, S.B., Islam, M.T., Al Maadeed, S., Zughaier, S.M., Khan, M.S., Chowdhury, M.E.: Exploring the effect of image enhancement techniques on covid-19 detection using chest x-ray images. Computers in Biol...

    2021

  27. [35]

    In: Advances in Neural Information Processing Systems (NeurIPS)

    Romano, Y., Sesia, M., Candes, E.: Classification with valid and adaptive coverage. In: Advances in Neural Information Processing Systems (NeurIPS). vol. 33, pp. 3581–3591 (2020) 2, 3, 7

    2020

  28. [36]

    In: International Joint Conference on Artificial Intelligence (IJCAI)

    Saunders, C., Gammerman, A., Vovk, V.: Transduction with confidence and cred- ibility. In: International Joint Conference on Artificial Intelligence (IJCAI). pp. 722–726 (1999) 2, 5, 6

    1999

  29. [37]

    In: Medical Image Com- puting and Computer-Assisted Intervention (MICCAI)

    Shakeri,F.,Huang,Y.,Silva-Rodríguez,J.,Bahig,H.,Tang,A.,Dolz,J.,BenAyed, I.: Few-shot adaptation of medical vision-language models. In: Medical Image Com- puting and Computer-Assisted Intervention (MICCAI). pp. 553–563 (2024) 1, 3, 8

    2024

  30. [38]

    Computer methods and programs in biomedicine195, 105637 (2020) 7

    Silva-Rodríguez, J., Colomer, A., Sales, M.A., Molina, R., Naranjo, V.: Going deeper through the gleason scoring scale: An automatic end-to-end system for histology prostate grading and cribriform pattern detection. Computer methods and programs in biomedicine195, 105637 (2020) 7

    2020

  31. [39]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Silva-Rodríguez, J., Hajimiri, S., Ayed, I.B., Dolz, J.: A closer look at the few-shot adaptation of large vision-language models. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp. 23681– 23690 (2024) 1, 2, 3, 5, 6, 8, 10, 11, 12

    2024

  32. [40]

    Medical Image Analysis99, 103357 (2025) 1, 7

    Silva-Rodríguez, J., Chakor, H., Kobbi, R., Dolz, J., Ayed, I.B.: A foundation language-image model of the retina (flair): Encoding expert knowledge in text supervision. Medical Image Analysis99, 103357 (2025) 1, 7

    2025

  33. [41]

    Stutz, D., Dvijotham, K.D., Cemgil, A.T., Doucet, A.: Learning optimal conformal classifiers.In:InternationalConferenceonLearningRepresentations(ICLR)(2022) 3, 4, 8

    2022

  34. [42]

    zero-shot

    Udandarao, V., Prabhu, A., Ghosh, A., Sharma, Y., Torr, P.H.S., Bibi, A., Al- banie, S., Bethge, M.: No "zero-shot" without exponential data: Pretraining con- cept frequency determines multimodal model performance. In: Advances in Neural Information Processing Systems (NeurIPS...

    2024

  35. [43]

    In: Proceedings of the Asian Conference on Machine Learning

    Vovk, V.: Conditional validity of inductive conformal predictors. In: Proceedings of the Asian Conference on Machine Learning. vol. 25, pp. 475–490 (2012) 2

    2012

  36. [44]

    Vovk,V.:Transductiveconformalpredictors.In:ArtificialIntelligenceApplications and Innovations. pp. 348–360 (2013) 2 16 J. Silva-Rodríguez et al

    2013

  37. [45]

    Springer (01 2005) 2, 4, 5

    Vovk, V., Gammerman, A., Shafer, G.: Algorithmic Learning in a Random World. Springer (01 2005) 2, 4, 5

    2005

  38. [46]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Wang, X., Peng, Y., Lu, L., Lu, Z., Bagheri, M., Summers, R.: Chestx-ray8: Hospital-scale chest x-ray database and benchmarks on weakly-supervised clas- sification and localization of common thorax diseases. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pat...

    2017

  39. [47]

    In: arXiv preprint arXiv:1911.04623 (2019) 8, 11, 12

    Wang, Y., Chao, W.L., Weinberger, K.Q., van der Maaten, L.: Simpleshot: Re- visiting nearest-neighbor classification for few-shot learning. In: arXiv preprint arXiv:1911.04623 (2019) 8, 11, 12

    2019 arXiv

  40. [48]

    In: Empirical Methods in Natural Language Pro- cessing (EMNLP)

    Wang, Z., Wu, Z., Agarwal, D., Sun, J.: Medclip: Contrastive learning from un- paired medical images and text. In: Empirical Methods in Natural Language Pro- cessing (EMNLP). pp. 1–12 (10 2022) 1, 7

    2022

  41. [49]

    In: Inter- national Conference on Machine Learning

    Xu, C., Xie, Y.: Conformal prediction interval for dynamic time-series. In: Inter- national Conference on Machine Learning. pp. 11559–11569. PMLR (2021) 2

    2021

  42. [50]

    In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Yu, T., Lu, Z., Jin, X., Chen, Z., Wang, X.: Task residual for tuning vision-language models. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). pp. 10899–10909 (2023) 3, 6

    2023

  43. [51]

    In: Euro- pean Conference on Computer Vision (ECCV)

    Zhang, R., Fang, R., Zhang, W., Gao, P., Li, K., Dai, J., Qiao, Y., Li, H.: Tip- adapter: Training-free clip-adapter for better vision-language modeling. In: Euro- pean Conference on Computer Vision (ECCV). pp. 1–19 (11 2022) 3, 8, 11, 12

    2022

  44. [52]

    In: Machine Learning for Healthcare (MHLC)

    Zhang, Y., Jiang, H., Miura, Y., Manning, C.D., Langlotz, C.P.: Contrastive learn- ing of medical visual representations from paired images and text. In: Machine Learning for Healthcare (MHLC). pp. 1–24 (2022) 7

    2022

  45. [53]

    International Journal of Computer Vision (2022) 3

    Zhou, K., Yang, J., Loy, C.C., Liu, Z.: Learning to prompt for vision-language models. International Journal of Computer Vision (2022) 3

    2022

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.