Pith. sign in

REVIEW 4 major objections 6 minor 2 cited by

Towards Interpretable Radiology Report Generation via Concept Bottlenecks using a Multi-Agentic RAG

T0 review · 4 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 Concept bottlenecks let a chest X-ray AI show its clinical reasoning while keeping 81% classification accuracy.

desk verdict A legitimate CBM + multi-agent RAG integration for chest X-ray reports with solid concept-intervention evidence, but the clinical-relevance claim rests entirely on LLM-as-a-judge scores against an AI-generated reference. read the letter →

arxiv 2412.16086 v2 pith:72DYYZYX submitted 2024-12-20 cs.IR cs.AIcs.CLcs.CVeess.IV

Hasan Md Tusfiqur Alam , Devansh Srivastav , Md Abdul Kadir , Daniel Sonntag This is my paper · ORCID
classification cs.IRcs.AIcs.CLcs.CVeess.IV
keywords interpretableradiologyreportgenerationconceptbottleneckmodelsmulti-agentretrieval-augmentedchestX-rayclassificationLLM-as-a-judgeCOVID-QUdatasetexplainableAIclinicaldecisionsupport
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 a chest X-ray classifier can remain highly accurate while exposing its reasoning, and that the exposed reasoning can be reused to write explainable radiology reports. Instead of classifying an image directly, the model first scores 20 clinical concepts per disease, normalizes these scores into an interpretable concept vector, and only then predicts the disease. That vector is passed to a team of three AI agents — a retrieval agent, a radiologist agent, and a medical writer agent — which generate the report. The authors report 81% classification accuracy on the COVID-QU dataset, above several baselines, and report-generation quality scores between 84% and 90% as judged by five language models. The practical interest is that correct behavior and inspectable, clinically meaningful explanations come from the same mechanism, not from a post-hoc overlay.

What carries the argument

The load-bearing object is the concept vector $e = (s_1, \dots, s_N)$, where each $s_i = \max(M_i)$ comes from max-pooling a cosine-similarity matrix between the image embedding from a chest X-ray vision-language model and the text embedding of a clinical concept. A linear layer maps the normalized vector to the class, so the classification is a transparent function of 20 named concepts. The report-generation side uses the same vector as input to a cascade of ReAct-style agents (agents that interleave reasoning with retrieval actions): a disease-specific retrieval agent pulls relevant clinical documents, a radiologist agent computes an influence score for each concept, and a medical writer agent composes the final report. The paper's argument that the pipeline is interpretable rests on this vector being both the classifier's only input and the report's organizing information.

What would settle it

Have board-certified radiologists rate a blinded set of multi-agent reports, single-agent reports, and GPT-4 reports for accuracy and clinical usefulness; if their ratings do not favor the multi-agent reports, the central claim of enhanced clinical relevance is unsupported. A complementary test would inspect whether the top-contributing concepts for correctly classified COVID-QU images correspond to the actual visible findings on the X-ray.

Watch

Extended reading notes

Core claim

The central discovery claim is that concept bottlenecks built from automatically discovered clinical concepts are not just a transparency add-on; they carry the predictive signal. For each image, embeddings from a chest X-ray vision-language model are compared by cosine similarity with embeddings of 20 medical descriptors per disease, max-pooled and normalized to produce a concept vector, and a single linear layer maps that vector to the disease class. The paper reports that this reaches 81% accuracy on COVID-QU, exceeding CLIP (47%), Bio-VIL (78%), Label-free CBM (72%), and Robust CBM (78%). It further shows that intervening on 3–4 top-contribution concepts for misclassified examples substantially improves accuracy, and that removing high-contribution concepts degrades performance sharply, both taken as evidence that the model genuinely reasons through concepts. The same concept vectors then drive a multi-agent retrieval-augmented generation cascade whose generated reports are rated higher than single-agent and GPT-4 reports by five LLM judges across semantic similarity, accuracy, correctness, clinical usefulness, and consistency.

Load-bearing premise

The evaluation assumes that five language models judging reports against an AI-written reference report give scores that match what a radiologist would consider accurate and clinically useful, and no radiologist or real clinical outcome is used in this paper.

Editorial extensions

If this is right

  • Concept bottleneck classifiers can match or exceed black-box baselines, so interpretability does not necessarily come at an accuracy cost on this dataset.
  • Because the concept vector is the classifier's only input, a clinician could correct a misclassified case by editing a few concept scores and watching both the diagnosis and the generated report change.
  • Routing report generation through separate retrieval, interpretation, and writing agents yields higher LLM-judged quality than a single agent or GPT-4 alone.
  • The poorer clustering metrics of the multi-agent reports are presented as a feature: COVID-19 and pneumonia reports sit closer because the biology overlaps, which a purely separation-driven metric would miss.

Reading between the lines

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

  • If concept vectors are as faithful as the intervention experiments suggest, the same vectors could be projected back onto the image to produce pixel-level localization of each contributing finding, a test the authors do not run.
  • Because concepts are discovered by GPT-4 and reports are judged by LLMs, the pipeline's explainability inherits whatever systematic blind spots those models have; expert-annotated concept labels on a small subset would ground the concept layer independently.
  • The 20-concept set is fixed per disease; dynamically selecting concepts for each image could improve both accuracy and the relevance of generated reports, but the paper does not test this.
  • One could make the reports more auditable by forcing the writer agent to cite which concept scores support each sentence, turning the generated text into a traceable argument rather than a free-form narrative.
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

4 major / 6 minor

Summary. The paper proposes a two-stage interpretable chest X-ray analysis framework: first, a concept bottleneck model (CBM) that uses ChexAgent image embeddings and GPT-4-generated clinical concepts to classify COVID-QU images and output concept contribution scores; second, a multi-agent retrieval-augmented generation (RAG) system with ReAct, Radiologist, and Medical Writer agents that produces radiology reports conditioned on the predicted class and concept scores. The authors report 81% classification accuracy on COVID-QU and evaluate generated reports via LLM-as-a-judge scores against a Dragonfly-Med reference, reporting five metrics between 84% and 90%, plus clustering metrics and a Mixture-of-Agents qualitative check. The central claims are that the concept bottleneck provides interpretability and that the multi-agent RAG enhances clinical relevance, explainability, and transparency of generated reports.

Significance. If the claims were supported, the combination of a concept bottleneck with a multi-agent RAG pipeline would be a useful contribution to interpretable medical report generation, and the concept-intervention and concept-removal experiments in Fig. 3 are a reasonable sanity check for the CBM component. The authors also provide a code repository, which aids reproducibility. However, the load-bearing evidence for the report-generation claim is currently self-referential: the 'ground-truth reference' is generated by another LLM, the judges are LLMs, and the MoA validation is also LLM-based, with no radiologist assessment or clinically validated metrics. The clustering results in Table 2 are worse for the multi-agent method on all reported indices and are reinterpreted post hoc. The classification result of 81% lacks confidence intervals and split details. The significance of the paper is therefore conditional on external validation that is not currently present.

major comments (4)
  1. [Section 3, Table 3] The report-generation evaluation uses LLM-as-a-judge scores against a 'ground-truth reference generated by Dragonfly-Med' (Section 3, para. 4), i.e., one LLM's output serves as the reference for other LLMs. With no radiologist ratings and no clinical metrics (e.g., factual correctness against expert annotations, RadGraph, or CheXpert-style labels), the abstract's claim that generated reports have 'clinical relevance' is unsupported. The five judge LLMs share language-model priors with the systems being evaluated, so the 84-90% scores may reflect stylistic self-consistency rather than clinical validity. The additional MoA check also uses LLM proposers and an LLM binary classifier, so it does not break the evaluative loop. Please add human expert evaluation or validated clinical metrics, with inter-rater agreement and statistical significance tests.
  2. [Section 3, Table 2 and Fig. 4] The Multi-Agent method is worse than both baselines on all four clustering indices (Silhouette 0.27 vs 0.41, Davies-Bouldin 1.44 vs 0.96, Calinski-Harabasz 44.78 vs 93.99, Dunn 0.36 vs 0.73), and the paper reinterprets this as 'reflecting clinical reality more accurately' because COVID-19 and Pneumonia overlap biologically. Without a quantitative clinical criterion, lower cluster separation does not provide evidence of better report quality; this is a post hoc explanation rather than a validated outcome. Please report a direct comparison with expert assessment or a pre-specified hypothesis for the expected cluster geometry.
  3. [Section 2.1, Table 1] The headline classification result of 81% accuracy on COVID-QU is reported without confidence intervals, number of runs, or a description of the train/validation/test split. In addition, the concept set C is generated by GPT-4 (Section 2.1), but no human validation or inter-rater agreement is reported for the concepts, and the concept set is not listed in the paper. The concept-removal and intervention experiments in Fig. 3 are useful sanity checks, but the paper should report means and standard deviations over multiple runs and the exact data partition before the 81% figure can be assessed.
  4. [Section 2.2] The retrieval relevance threshold τ in the definition of D_C is a free parameter that directly controls which documents the ReAct agent retrieves, yet its value is not reported and no sensitivity analysis is given. The clinical document collection is described only as 'clinical documentation for each disease category from the NIH'; its size, curation, and distribution across categories are unspecified. Without these details the report-generation pipeline is not reproducible, and the effect of retrieval on the Table 3 scores cannot be isolated.
minor comments (6)
  1. [Table 1] 'Comparision' and 'Intepretability' are misspelled, and 'Covid-QU' should be 'COVID-QU' consistently throughout the paper.
  2. [Table 1] Bio-VIL is cited as [37] in the table, but the text in Section 3 introduces it as [2]; please correct the citation.
  3. [Section 3] The MoA evaluation says 'As shown in the table below' but no table is included; the scores 0.81/0.82/0.85 appear only in the prose.
  4. [Table 3] There are formatting/spacing issues in model names ('LLaV A 9B', 'GPT 3.5 T urbo') that should be fixed.
  5. [Section 2.1] The notation for the similarity matrix and max pooling is ambiguous: M_{i,j} ∈ R^{H×N} followed by s_i = max(M_i) should clarify whether i indexes spatial locations, patches, or samples.
  6. [Abstract and Section 4] The abstract states the code is available at the GitHub URL, while the full text says 'code will be released'; please reconcile the two statements.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the classification and report-generation pipeline is forward and externally benchmarked, while the LLM-as-judge evaluation loop is an external-validity limitation rather than a circular derivation.

full rationale

The paper's claimed derivation chain is not equivalent to its inputs by construction. Classification is trained on the externally labeled COVID-QU dataset (33,920 images with COVID-19/Pneumonia/Normal labels); the concept set is obtained from GPT-4 prompts following [37], image embeddings from ChexAgent, and a cosine-similarity max-pooled concept vector feeds a linear classifier trained with cross-entropy. Accuracy (81%) is measured against human disease labels and compared with independent baselines (CLIP, Bio-VIL, Label-free CBM, Robust CBM), so the central classification claim is not fitted to the evaluation target. The concept-removal and concept-intervention experiments further test the model's dependence on concept contributions rather than assuming it. Report generation is a forward pipeline: the predicted class and concept contribution scores condition the multi-agent RAG system, and the generated reports are clustered and judged. The only evidence for 'clinical relevance' is the LLM-as-a-judge evaluation in Table 3, where five LLMs score reports against a 'ground-truth reference generated by Dragonfly-Med,' itself an LLM. This is a genuine external-validation gap—the scores may reflect LLM text similarity rather than radiologist utility—but it is not a circular step in the derivation: the report text is not constructed from the reference, no parameter is fitted to the judge scores, and no equation reduces an output to an input. The two author-overlapping references ([22], [31]) appear only as incidental related work and do not carry the argument. Therefore no circularity score is warranted.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new scientific entities; its load-bearing choices are the hand-picked concept set, the retrieval threshold, and the assumption that LLM-based evaluation and GPT-generated concepts are clinically valid.

free parameters (2)
  • Concept set C (20 concepts per disease) = 20 concepts per disease for COVID-19, Pneumonia, Normal, generated by GPT-4
    The concept descriptors are chosen by prompting GPT-4, per Section 2.1. They are not learned from data and are a design choice that directly shapes the concept bottleneck.
  • Retrieval relevance threshold tau = not specified
    The threshold tau in the retrieval step (D_C = {e_j | sim(v_j,q) >= tau}) is a free parameter that determines which documents are retrieved; its value is not reported.
assumptions (4)
  • domain assumption COVID-QU dataset labels are correct
    Classification accuracy is measured against COVID-QU labels with no error analysis or human re-check.
  • domain assumption LLM-as-a-judge scores correlate with clinical utility
    The report generation evaluation assumes that five LLMs' judgments on Semantic Similarity, Accuracy, Correctness, Clinical Usefulness, and Consistency are a valid proxy for what clinicians need; no human validation is provided.
  • domain assumption ChexAgent embeddings capture the visual features relevant to the disease classes
    The concept bottleneck relies on ChexAgent image embeddings, taken as a fixed feature extractor; if these embeddings miss clinically relevant findings, the concepts and reports built on them are incomplete.
  • domain assumption GPT-4 generated concepts are the clinically relevant set
    The 20 concepts per disease are produced by GPT-4 prompts and are not validated against expert-defined concepts; the choice of concepts determines the interpretability of the bottleneck.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Interpretable Radiology Report Generation via Concept Bottlenecks using a Multi-Agentic RAG." pith.science (2026). https://pith.science/paper/72DYYZYX

@misc{pith2026241216086,
  author       = {Pith},
  title        = {Pith review of: Towards Interpretable Radiology Report Generation via Concept Bottlenecks using a Multi-Agentic RAG},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/72DYYZYX}},
  note         = {Machine review of arXiv:2412.16086}
}
read the original abstract

Deep learning has advanced medical image classification, but interpretability challenges hinder its clinical adoption. This study enhances interpretability in Chest X-ray (CXR) classification by using concept bottleneck models (CBMs) and a multi-agent Retrieval-Augmented Generation (RAG) system for report generation. By modeling relationships between visual features and clinical concepts, we create interpretable concept vectors that guide a multi-agent RAG system to generate radiology reports, enhancing clinical relevance, explainability, and transparency. Evaluation of the generated reports using an LLM-as-a-judge confirmed the interpretability and clinical utility of our model's outputs. On the COVID-QU dataset, our model achieved 81% classification accuracy and demonstrated robust report generation performance, with five key metrics ranging between 84% and 90%. This interpretable multi-agent framework bridges the gap between high-performance AI and the explainability required for reliable AI-driven CXR analysis in clinical settings. Our code is available at https://github.com/tifat58/IRR-with-CBM-RAG.git.

Figures

Figures reproduced from arXiv: 2412.16086 by the authors.

Figure 1
Figure 1. Input (blue) to output (green) pipeline: Given a CXR as input, the Con￾cept Bottleneck Model predicts clinical attributes (concepts) and their contri￾butions in an intermediate step, followed by predicting the disease class. The multi-agent RAG system then generates a comprehensive report, incorporating clinical interpretations and insights drawn from relevant clinical documents concept ci , a text embedding ti ∈ R … view at source ↗
Figure 2
Figure 2. Proposed architecture for the interpretable report generation. ( [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Evaluation of the Robustness of Concept set of the classification model. [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: t-SNE visualization of the embeddings of generated reports [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. Engineering RAG Systems for Real-World Applications: Design, Development, and Evaluation

    cs.SE 2025-06 conditional novelty 4.0 of 10

    A field study of five real-world RAG systems evaluated by 100 users, yielding user ratings and twelve engineering lessons.

  2. Towards Automated Situation Awareness: A RAG-Based Framework for Peacebuilding Reports

    cs.CY 2025-05 conditional novelty 4.0 of 10

    A dynamic RAG pipeline generates situation awareness reports for peacebuilding from GDELT, ACLED, ReliefWeb, and World Bank data, evaluated by NLP metrics, UNDP experts, and LLM judges.

Reference graph

Works this paper leans on

40 extracted references · 9 canonical work pages · cited by 2 Pith papers

  1. [1]

    arXiv preprint arXiv:2303.08774 (2023)

    Achiam, J., Adler, S., Agarwal, S., Ahmad, L., Akkaya, I., Aleman, F.L., Almeida, D., Altenschmidt, J., Altman, S., Anadkat, S., et al.: Gpt-4 technical report. arXiv preprint arXiv:2303.08774 (2023)

    arXiv 2023

  2. [2]

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

    Bannur, S., Hyland, S., Liu, Q., Perez-Garcia, F., Ilse, M., Castro, D.C., Boecking, B., Sharma, H., Bouzid, K., Thieme, A., et al.: Learning to exploit temporal struc- ture for biomedical vision-language processing. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 15016–15027 (2023)

    2023

  3. [3]

    In: 2024 International Joint Conference on Neural Networks (IJCNN)

    Bernardi, M.L., Cimitile, M.: Report generation from x-ray imaging by retrieval- augmented generation and improved image-text matching. In: 2024 International Joint Conference on Neural Networks (IJCNN). pp. 1–8. IEEE (2024)

    work page 2024

  4. [4]

    Bishop, C.M., Nasrabadi, N.M.: Pattern recognition and machine learning, vol. 4. Springer (2006)

    2006

  5. [5]

    arXiv preprint arXiv:2005.14165 (2020)

    Brown, T.B.: Language models are few-shot learners. arXiv preprint arXiv:2005.14165 (2020)

    arXiv 2020

  6. [6]

    Communications in Statistics-theory and Methods3(1), 1–27 (1974)

    Caliński, T., Harabasz, J.: A dendrite method for cluster analysis. Communications in Statistics-theory and Methods3(1), 1–27 (1974)

    1974

  7. [7]

    In: Proceedings of the AAAI Conference on Artificial In- telligence

    Chauhan, K., Tiwari, R., Freyberg, J., Shenoy, P., Dvijotham, K.: Interactive con- cept bottleneck models. In: Proceedings of the AAAI Conference on Artificial In- telligence. vol. 37, pp. 5948–5955 (2023)

    work page 2023

  8. [8]

    arXiv preprint arXiv:2406.00977 (2024)

    Chen, K., Thapa, R., Chalamala, R., Athiwaratkun, B., Song, S.L., Zou, J.: Dragonfly: Multi-resolution zoom supercharges large visual-language model. arXiv preprint arXiv:2406.00977 (2024)

    arXiv 2024

Show all 40 references
  1. [9]

    arXiv preprint arXiv:2401.12208 (2024)

    Chen, Z., Varma, M., Delbrouck, J.B., Paschali, M., Blankemeier, L., Van Veen, D., Valanarasu, J.M.J., Youssef, A., Cohen, J.P., Reis, E.P., et al.: Chexagent: Towards a foundation model for chest x-ray interpretation. arXiv preprint arXiv:2401.12208 (2024)

    2024 arXiv

  2. [10]

    Chowdhury, M.E., Rahman, T., Khandakar, A., Mazhar, R., Kadir, M.A., Mahbub, Z.B., Islam, K.R., Khan, M.S., Iqbal, A., Al Emadi, N., et al.: Can ai help in screening viral and covid-19 pneumonia? Ieee Access8, 132665–132676 (2020)

    2020

  3. [11]

    IEEE transactions on pattern analysis and machine intelligence (2), 224–227 (1979)

    Davies, D.L., Bouldin, D.W.: A cluster separation measure. IEEE transactions on pattern analysis and machine intelligence (2), 224–227 (1979)

    1979

  4. [12]

    Neural Computing and Applications35(22), 16061–16071 (2023)

    De Falco, I., De Pietro, G., Sannino, G.: Classification of covid-19 chest x-ray images by means of an interpretable evolutionary rule-based approach. Neural Computing and Applications35(22), 16061–16071 (2023)

    2023

  5. [13]

    arXiv preprint arXiv:2407.21783 (2024)

    Dubey, A., Jauhri, A., Pandey, A., Kadian, A., Al-Dahle, A., Letman, A., Mathur, A., Schelten, A., Yang, A., Fan, A., et al.: The llama 3 herd of models. arXiv preprint arXiv:2407.21783 (2024)

    2024 arXiv

  6. [14]

    Journal of cy- bernetics 4(1), 95–104 (1974)

    Dunn, J.C.: Well-separated clusters and optimal fuzzy partitions. Journal of cy- bernetics 4(1), 95–104 (1974)

    1974

  7. [15]

    Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery9(4), e1312 (2019)

    Holzinger, A., Langs, G., Denk, H., Zatloukal, K., Müller, H.: Causability and explainability of artificial intelligence in medicine. Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery9(4), e1312 (2019)

    2019

  8. [16]

    JAMA network open2(3), e191095–e191095 (2019)

    Hwang, E.J., Park, S., Jin, K.N., Im Kim, J., Choi, S.Y., Lee, J.H., Goo, J.M., Aum, J., Yim, J.J., Cohen, J.G., et al.: Development and validation of a deep learning–based automated detection algorithm for major thoracic diseases on chest radiographs. JAMA network open2(3), e...

    2019

  9. [17]

    arXiv preprint arXiv:2310.06825 (2023) Towards Interpretable Radiology Report

    Jiang, A.Q., Sablayrolles, A., Mensch, A., Bamford, C., Chaplot, D.S., Casas, D.d.l., Bressand, F., Lengyel, G., Lample, G., Saulnier, L., et al.: Mistral 7b. arXiv preprint arXiv:2310.06825 (2023) Towards Interpretable Radiology Report ... 9

    2023 arXiv

  10. [18]

    arXiv preprint arXiv:1711.08195 (2017)

    Jing, B., Xie, P., Xing, E.: On the automatic generation of medical imaging reports. arXiv preprint arXiv:1711.08195 (2017)

    2017 arXiv

  11. [19]

    In: International conference on machine learning

    Koh, P.W., Nguyen, T., Tang, Y.S., Mussmann, S., Pierson, E., Kim, B., Liang, P.: Concept bottleneck models. In: International conference on machine learning. pp. 5338–5348. PMLR (2020)

    2020

  12. [20]

    Advances in Neural Information Processing Systems 33, 9459–9474 (2020)

    Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., Küttler, H., Lewis, M., Yih, W.t., Rocktäschel, T., et al.: Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems 33, 9459–9474 (2020)

    2020

  13. [21]

    arXiv preprint arXiv:2402.05120 (2024)

    Li, J., Zhang, Q., Yu, Y., Fu, Q., Ye, D.: More agents is all you need. arXiv preprint arXiv:2402.05120 (2024)

    2024 arXiv

  14. [22]

    In: Proceedings of the 1st Workshop on Uncertainty-Aware NLP (UncertaiNLP 2024)

    Liang, S., Sánchez, P., Sonntag, D.: Optimizing relation extraction in medical texts through active learning: A comparative analysis of trade-offs. In: Proceedings of the 1st Workshop on Uncertainty-Aware NLP (UncertaiNLP 2024). pp. 23–34 (2024)

    2024

  15. [23]

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

    Liu, H., Li, C., Li, Y., Lee, Y.J.: Improved baselines with visual instruction tuning. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. pp. 26296–26306 (2024)

    2024

  16. [24]

    Journal of machine learning research 9(11) (2008)

    Van der Maaten, L., Hinton, G.: Visualizing data using t-sne. Journal of machine learning research 9(11) (2008)

    2008

  17. [25]

    arXiv preprint arXiv:2304.06129 (2023)

    Oikarinen, T., Das, S., Nguyen, L.M., Weng, T.W.: Label-free concept bottleneck models. arXiv preprint arXiv:2304.06129 (2023)

    2023 arXiv

  18. [26]

    In: International conference on machine learning

    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. pp. 8748–8763. PMLR (2021)

    2021

  19. [27]

    In: Machine Learning for Healthcare Conference

    Ranjit, M., Ganapathy, G., Manuel, R., Ganu, T.: Retrieval augmented chest x-ray report generation using openai gpt models. In: Machine Learning for Healthcare Conference. pp. 650–666. PMLR (2023)

    2023

  20. [28]

    Journal of computational and applied mathematics20, 53–65 (1987)

    Rousseeuw, P.J.: Silhouettes: a graphical aid to the interpretation and validation of cluster analysis. Journal of computational and applied mathematics20, 53–65 (1987)

    1987

  21. [29]

    In: International Conference on Machine Learning

    Shin, S., Jo, Y., Ahn, S., Lee, N.: A closer look at the intervention procedure of concept bottleneck models. In: International Conference on Machine Learning. pp. 31504–31520. PMLR (2023)

    2023

  22. [30]

    Nature 620(7972), 172–180 (2023)

    Singhal, K., Azizi, S., Tu, T., Mahdavi, S.S., Wei, J., Chung, H.W., Scales, N., Tanwani, A., Cole-Lewis, H., Pfohl, S., et al.: Large language models encode clinical knowledge. Nature 620(7972), 172–180 (2023)

    2023

  23. [31]

    In: 2021 11th international conference on cloud computing, data science & engineering (confluence)

    Srivastav, D., Bajpai, A., Srivastava, P.: Improved classification for pneumonia detection using transfer learning with gan based synthetic image augmentation. In: 2021 11th international conference on cloud computing, data science & engineering (confluence). pp. 433–437. IEEE (2021)

    2021

  24. [32]

    arXiv preprint arXiv:2308.13453 (2023)

    Steinmann, D., Stammer, W., Friedrich, F., Kersting, K.: Learning to intervene on concept bottlenecks. arXiv preprint arXiv:2308.13453 (2023)

    2023 arXiv

  25. [33]

    arXiv preprint arXiv:2407.15268 (2024)

    Sun, L., Zhao, J., Han, M., Xiong, C.: Fact-aware multimodal retrieval aug- mentation for accurate medical radiology report generation. arXiv preprint arXiv:2407.15268 (2024)

    2024 arXiv

  26. [34]

    arXiv preprint arXiv:2408.00118 (2024)

    Team, G., Riviere, M., Pathak, S., Sessa, P.G., Hardin, C., Bhupatiraju, S., Hussenot, L., Mesnard, T., Shahriari, B., Ramé, A., et al.: Gemma 2: Improving open language models at a practical size. arXiv preprint arXiv:2408.00118 (2024)

    2024 arXiv

  27. [35]

    arXiv preprint arXiv:2406.04692 (2024) 10 Alam et al

    Wang, J., Wang, J., Athiwaratkun, B., Zhang, C., Zou, J.: Mixture-of-agents en- hances large language model capabilities. arXiv preprint arXiv:2406.04692 (2024) 10 Alam et al

    2024 arXiv

  28. [36]

    arXiv preprint arXiv:2410.13085 (2024)

    Xia, P., Zhu, K., Li, H., Wang, T., Shi, W., Wang, S., Zhang, L., Zou, J., Yao, H.: Mmed-rag: Versatile multimodal rag system for medical vision language models. arXiv preprint arXiv:2410.13085 (2024)

    2024 arXiv

  29. [37]

    arXiv preprint arXiv:2310.03182 (2023)

    Yan, A., Wang, Y., Zhong, Y., He, Z., Karypis, P., Wang, Z., Dong, C., Gentili, A., Hsu, C.N., Shang, J., et al.: Robust and interpretable medical image classifiers via concept bottleneck models. arXiv preprint arXiv:2310.03182 (2023)

    2023 arXiv

  30. [38]

    arXiv preprint arXiv:2210.03629 (2022)

    Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K., Cao, Y.: React: Syn- ergizing reasoning and acting in language models. arXiv preprint arXiv:2210.03629 (2022)

    2022 arXiv

  31. [39]

    arXiv preprint arXiv:2205.15480 (2022)

    Yuksekgonul, M., Wang, M., Zou, J.: Post-hoc concept bottleneck models. arXiv preprint arXiv:2205.15480 (2022)

    2022 arXiv

  32. [40]

    Advances in Neural Information Processing Systems36, 46595–46623 (2023)

    Zheng, L., Chiang, W.L., Sheng, Y., Zhuang, S., Wu, Z., Zhuang, Y., Lin, Z., Li, Z., Li, D., Xing, E., et al.: Judging llm-as-a-judge with mt-bench and chatbot arena. Advances in Neural Information Processing Systems36, 46595–46623 (2023)

    2023

Pith tools

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