Pith. sign in

REVIEW 3 major objections 6 minor 17 references

Agentic Visual Reasoning in Whole-Slide Pathology Images via Active Perception

T0 review · 3 major / 6 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read This paper claims that a learned, abnormality-driven active-perception Navigator makes zero-shot reasoning over gigapixel pathology slides state-of-the-art and traceable.

desk verdict A well-evaluated active-perception WSI system with a genuinely pathology-grounded Navigator; the main open question is how much the question-agnostic abnormality ranking limits generalization to evidence types it wasn't trained to flag. read the letter →

arxiv 2608.08648 v1 pith:RALEEP2J submitted 2026-08-09 cs.CV

Jingyun Chen , Fengchun Liu , Linghan Cai , Songhan Jiang , Shenjin Huang , Hongpeng Wang , Lequan Yu , Yongbing Zhang This is my paper · ORCID
classification cs.CV
keywords whole-slidepathologyactiveperceptionvisualquestionansweringreinforcementlearningmulti-magnificationnavigationcancersubtypeclassificationevidencedeliberationzero-shotreasoning
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 argues that the right way to reason over a gigapixel pathology slide is neither to compress the whole image into one representation nor to wander heuristically, but to actively decide where to look, at what zoom, and in what order. It proposes AdaptivePath, whose Navigator is trained on pathologist-reviewed normal/abnormal labels to choose observation locations and spatial extents, and whose Interpreter, Deliberator, and Arbiter convert those observations into question-conditioned evidence, option-by-option audits, and a final answer. The paper reports state-of-the-art zero-shot accuracy on slide-level and region-level pathology visual question answering, 80.14 percent accuracy for cancer subtype classification across six cohorts, and 82.9 percent accuracy when pathologists diagnose from AdaptivePath-selected image sequences. If these results hold, agentic visual reasoning over whole slides becomes both practical and traceable without costly question-specific trajectory annotations.

What carries the argument

The load-bearing object is the Navigator, a two-part policy consisting of FocusNet, which encodes a global glance and the sequence of focus regions into a visual state, and ActionNet, which predicts abnormality logits and the next focus action (location plus relative spatial extent). It is trained in two phases: first alternating representation learning with proximal policy optimization (a standard clipped reinforcement-learning update), using as reward the reduction in abnormality-classification loss from each observation; second, consistency-regularized fine-tuning that enforces the same focus trajectories under geometric and appearance-preserving transformations. The remaining components, the question-conditioned Morphology Interpreter, the option-wise Deliberator, and the cross-scale Arbiter, turn the acquired focus regions into explicit evidence audits. The mechanism works because the abnormality score acts as a cheap, question-agnostic proxy for where diagnostic evidence lives, and the hierarchical magnification schedule lets a small budget of regions cover sparse, scale-varying morphology.

What would settle it

Take a slide-level question whose correct answer depends on regions a pathologist would label normal rather than abnormal, such as a question about the absence of tumor in a specific compartment, and run AdaptivePath with its default budget of five regions per magnification and four focus observations per region. If the retained high-magnification regions contain none of the question-critical normal regions and accuracy collapses toward random guessing, while the same question answered from the full-slide context succeeds, then the claim that abnormality ranking suffices for evidence acquisition is refuted.

Watch

Extended reading notes

Core claim

AdaptivePath's central claim is that a pathology-grounded, question-agnostic navigation policy can replace exhaustive patch compression and heuristic region retrieval for whole-slide reasoning. The Navigator is trained to recognize abnormal regions and to output the location and relative spatial extent of each next observation, using pathologist-reviewed binary normal/abnormal labels rather than question-specific evidence annotations. At inference it works hierarchically from low to high magnification, retaining the top-ranked regions by an abnormality score at each level, so a limited observation budget is spent where diagnostic morphology actually lives. Each observed region is converted into a question-conditioned morphological description; a Deliberator then audits how the directly observed morphology supports or conflicts with every answer option and rates evidence sufficiency; and an Arbiter integrates these audits across magnifications into the final answer. The paper reports that this pipeline outperforms existing agentic systems on WSI and region pathology VQA benchmarks and raises zero-shot cancer subtype classification accuracy to 80.14 percent on the same cohorts.

Load-bearing premise

Everything rests on the assumption that a binary normal/abnormal score, learned from pathologist-reviewed labels, is a sufficient proxy for where the evidence for any clinical question lives; if question-relevant evidence appears in tissue a pathologist would call normal, the Navigator can discard it before the reasoner ever sees it.

Editorial extensions

If this is right

  • Zero-shot slide-level visual question answering becomes practical: a frozen vision-language reasoner can answer questions over a whole slide accurately when it is fed a small set of actively chosen, multi-scale regions instead of a downsampled thumbnail or random patches.
  • Cancer subtype classification can be performed from H&E slides alone, with no subtype labels used in training, indicating that the learned navigation transfers to tasks beyond the abnormality-recognition objective it was trained on.
  • Each final answer is retraceable: the system records which regions were observed, what morphology was directly seen, and which answer options were supported, contradicted, or left neutral.
  • The observation budget behaves predictably: retaining five regions per magnification outperforms both smaller and larger budgets, and no single magnification matches the full low-to-high sequence's diagnostic utility in the paper's pathologist study.
  • Because the system currently reads only H&E morphology, questions whose evidence requires immunohistochemistry or molecular assays remain outside its directly observable evidence, as the paper states in its conclusion.

Reading between the lines

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

  • Not tested in the paper: because the Navigator is question-agnostic, the same trained policy could be reused for other gigapixel-image tasks, such as finding morphologically similar cases or highlighting regions relevant to survival analysis, as long as the target evidence is visible as abnormality.
  • A testable gap implied by the design: if a question's correct answer depends on tissue a pathologist would call normal, such as the absence of invasive cells in a particular compartment, the abnormality-based ranking may spend the budget elsewhere; a small amount of question-aware reranking data could close that gap.
  • The consistency-regularization phase hints at a general recipe for stabilizing reinforcement-learned visual observation policies under stain and rotation variation, which could transfer to other gigapixel imaging domains such as satellite or materials imagery.
  • The 82.9 percent pathologist accuracy on AdaptivePath-selected sequences, compared with 89.7 percent on recorded pathologist trajectories, sets a concrete calibration target: closing that remaining gap is a measurable next milestone.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. This paper presents AdaptivePath, an agentic system for whole-slide pathology image reasoning. A Navigator trained on pathologist-reviewed abnormal-region labels uses PPO with alternating representation learning and consistency fine-tuning to decide where and at what magnification to acquire observations. At inference, candidate regions are hierarchically pruned using an abnormality score, and a frozen Patho-R1-7B Interpreter converts selected foci into question-conditioned morphology descriptions; a Qwen3-8B Deliberator and Arbiter audit evidence per option and integrate across magnifications. The paper reports state-of-the-art zero-shot results on SlideBench-VQA (60.81%), WSI-VQA (54.92%), and PathVQA (60.73%), 80.14% accuracy on six TCGA subtype classification cohorts, and a reader study in which pathologists achieve 82.9% accuracy using AdaptivePath-selected sequences. The central claim is that question-agnostic abnormality-driven active perception with explicit evidence deliberation provides an effective and traceable basis for WSI reasoning.

Significance. If correct, AdaptivePath is a meaningful step toward agentic WSI reasoning: it replaces heuristic region selection with a policy trained on pathologist-reviewed abnormality labels, supports multi-scale evidence acquisition under a fixed ROI budget, and makes the evidence chain explicit through interpretation, deliberation, and adjudication. The paper includes extensive comparisons, ablations of ROI selection, magnification, and budget, a two-configuration module study, and a blinded diagnostic utility test, and the code is released. The main strength is the combination of learned active perception with explicit evidence auditing, evaluated across both VQA and classification. However, the significance of the headline numbers depends on closing three gaps: confidence intervals around the benchmark differences, a demonstration that abnormality-only pruning retains question-relevant evidence for categories not aligned with generic abnormality (e.g., HER2 expression and molecular subtype), and clarity about how much of the navigation signal comes from the Patho-R1-7B-generated labels versus independent pathologist annotation.

major comments (3)
  1. [Section 3.3.1, Eqs. (7)-(8)] Question-agnostic abnormality ranking can discard question-relevant evidence before the Interpreter sees it. The Navigator is trained only on binary normal/abnormal labels (Section 4.1), and candidate retention is based solely on AS(Rm); no question information enters until after top-K pruning. For questions whose decisive morphology is not 'abnormal' in the training sense (e.g., HER2 expression or molecular subtype morphology in H&E, or negative/absence findings), the correct region may be removed at the 2.5x/5x stage. Table 1 is consistent with this failure mode: HER2 Expression and Molecular Subtype accuracies are 27.92 and 32.68, far below Tumor Type at 89.58. The paper also lists question-aware reranking as future work in Section 5, which concedes the limitation. I ask the authors to report evidence recall in retained regions per question category, or to compare against a question-aware reranker; otherwise the generalization claim for arbitrary clinical questions is not established.
  2. [Tables 1 and 2] No error bars, confidence intervals, or numbers of independent runs are reported for the main benchmark results. Several state-of-the-art claims rest on small margins (e.g., 1.39 points over PathNavigate on SlideBench-VQA and 2.71 points on WSI-VQA), and Table 3 ablations also lack variance. Without bootstrapped confidence intervals or multi-seed evaluation, these differences cannot be distinguished from noise. Since the paper already computes paired t-tests for Figure 4, the same uncertainty reporting should be applied to the headline tables.
  3. [Section 4.1] The abnormal-region labels are initially generated by Patho-R1-7B and later reviewed and corrected by pathologists, but the paper reports no label correction rate, inter-reviewer agreement, or proportion of labels changed. Because Patho-R1-7B is also the default Interpreter, the Navigator may be optimized toward the same model's notion of 'abnormal', and the end-to-end gain may partly reflect model self-consistency rather than independent pathology grounding. Please quantify the review process and include an ablation with labels from an independent source (e.g., manual annotation or a different model), or at least report results on a held-out set of fully manual labels.
minor comments (6)
  1. [Table 2] The 'Overall' column needs its weighting formula; the unweighted average of the six AdaptivePath cohort accuracies is 78.95%, not 80.14%, so either the cohorts are weighted by size or the overall value is computed differently.
  2. [Section 4.6.1, Figure 3] The reader study's sample size, number of pathologists, and selection criteria for the SlideBench-VQA subset are not stated in the main text; please add them or point to a specific supplementary section.
  3. [Section 3.2.3, Eq. (6)] Please define how the aggregate heatmap H is normalized before computing D_JS and state whether H_geo is mapped back to the original coordinate system before aggregation.
  4. [Section 4.4] The Patch baseline samples 30 patches at 20x randomly; without fixing seeds or averaging over repeated samples, the reported patch-model scores are stochastic. Please report the sampling procedure in more detail.
  5. [Throughout] There are several typographical issues: 'Y ongbing Zhang' in the author block, irregular spacing in 'Quilt-LLaV A' and 'LLaV A', and inconsistent use of 'Patho-R1-7B' versus 'Patho-R1-7B' in a few places. A careful proofread is recommended.
  6. [Section 4.3] PathMMU results are mentioned only as being in the Supplementary Materials; a one-sentence summary in the main text would help readers who do not consult the supplement.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the navigator is trained on pathologist-corrected abnormality labels and evaluated on external VQA/classification targets.

full rationale

AdaptivePath's derivation chain is not circular. The Navigator is trained exclusively on binary normal/abnormal labels (Section 4.1) that, although initially generated by Patho-R1-7B, are reviewed and corrected by six pathologists, and the policy objective (Eqs. 3-5) and inference ranking (Eqs. 7-8) never use the downstream VQA answers or subtype labels. The abnormality score AS(Rm) is a learned intermediate saliency measure, not a re-encoding of the final prediction; the final answer is produced by the frozen Deliberator/Arbiter from region morphology sets (Eqs. 9-13). The use of Patho-R1-7B both as label generator and as Interpreter creates a potential self-referential loop in evidence acquisition, but it does not force the reported results because the training labels are human-corrected and the benchmark answers are external. The paper explicitly acknowledges that H&E-only evidence cannot address IHC/molecular findings, which is a limitation rather than a circular dependency. No fitted parameter is renamed as a prediction, and no load-bearing self-citation chain is present; PathAgent is cited only as a baseline. The question-agnostic pruning in Eq. 7-8 may pose a correctness risk for evidence not marked 'abnormal', but this is an assumption about generalization, not circularity.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the reliability of pathologist-reviewed abnormality labels, the transferability of abnormality-driven navigation to arbitrary clinical questions, and the faithfulness of frozen LLM components. The free parameters are mostly inference budgets and training hyperparameters. No new physical or biological entities are introduced.

free parameters (5)
  • ROI budget K = 5
    Number of candidate regions retained at each magnification. Ablated in Table 3C, where K=5 is best; a design choice that affects the navigation budget.
  • Focus budget T = 4
    Number of focus observations per selected region. Stated as a default inference parameter, not ablated in the main paper.
  • Magnification set M = {2.5x, 5x, 10x}
    Hierarchical navigation magnifications. Ablated in Table 3B; combining all three gives the best results.
  • Consistency weights alpha, beta_stab, beta_ent = alpha=1, beta_stab=0.01, beta_ent=0.01
    Hyperparameters in Eqs. 3 and 5 controlling distillation, stability, and entropy regularization in the Navigator training.
  • Region input size P and focus size P_f = P=512, P_f=112
    Input resolutions for FocusNet and ActionNet. Stated as implementation detail, not individually ablated.
assumptions (4)
  • domain assumption Pathologist-reviewed abnormal-region labels provide reliable ground truth for navigation utility.
    The entire Navigator training relies on the binary normal/abnormal labels being meaningful for routing attention to diagnostically relevant regions. Entered in Section 4.1.
  • domain assumption The reward defined as reduction in classification loss incentivizes informative observations.
    Eq. 4 assumes that a focus which improves abnormality recognition is a good observation for downstream question answering. This is plausible but not proven for arbitrary clinical questions.
  • domain assumption The frozen Interpreter and Deliberator LLMs (Patho-R1-7B, Qwen3-8B) provide faithful morphological descriptions and evidence audits.
    All downstream reasoning is performed by these pretrained models; their reliability is taken as given, and no independent verification of their outputs is reported in the main text.
  • standard math Standard RL assumptions for PPO hold.
    The paper uses clipped PPO with GAE; standard convergence and stability assumptions apply without additional verification.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Agentic Visual Reasoning in Whole-Slide Pathology Images via Active Perception." pith.science (2026). https://pith.science/paper/RALEEP2J

@misc{pith2026260808648,
  author       = {Pith},
  title        = {Pith review of: Agentic Visual Reasoning in Whole-Slide Pathology Images via Active Perception},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RALEEP2J}},
  note         = {Machine review of arXiv:2608.08648}
}
read the original abstract

Whole-slide visual reasoning requires identifying sparse diagnostic evidence in gigapixel pathology slides and integrating observations across spatial scales. Existing WSI methods either compress densely sampled patches into global representations or use pretrained vision-language models with heuristic region selection, weakening links between predictions and morphology or lacking pathology-trained observation policies. We present AdaptivePath, an active-perception framework that formulates WSI evidence acquisition as sequential decision making. The Navigator learns question-agnostic abnormality-driven navigation from pathologist-reviewed labels to select observation locations and spatial extents, avoiding costly question-specific trajectory annotations. We train this policy through alternating representation learning and proximal policy optimization, followed by fine-tuning with geometric and appearance consistency objectives to stabilize focus trajectories. During inference, the Navigator hierarchically acquires sparse observations from low to high magnification under a limited ROI budget. A Morphology Interpreter converts observations into question-conditioned evidence, while the Deliberator evaluates evidence and revises intermediate answers across magnifications. The Arbiter integrates deliberation history to produce final answers. AdaptivePath achieves state-of-the-art zero-shot performance on WSI and region pathology VQA benchmarks and reaches 80.14% accuracy for cancer subtype classification across six TCGA cohorts. In a blinded diagnostic-utility study, pathologists using AdaptivePath-selected observation sequences achieve 82.9% accuracy. These results demonstrate that learned active perception enables effective and traceable visual reasoning over gigapixel pathology slides.

Figures

Figures reproduced from arXiv: 2608.08648 by the authors.

Figure 1
Figure 1. Overview of AdaptivePath. The Navigator is trained on abnormal-region recognition [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Overview of the active-perception Navigator. FocusNet encodes the global glance and [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Diagnostic utility test on a subset of SlideBench-VQA (BCNB) under the same ROI budget. [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Navigator ablations on WSI-VQA and cancer classification. Asterisks indicate significance [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Qualitative navigation comparison under the same ROI budget. The blue line and black [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 5 canonical work pages

  1. [5]

    Xuehai He, Yichen Zhang, Luntian Mou, Eric Xing, and Pengtao Xie

    URL https://arxiv.org/abs/2404.15127. Xuehai He, Yichen Zhang, Luntian Mou, Eric Xing, and Pengtao Xie. Pathvqa: 30000+ questions for medical visual question answering,

    arXiv

  2. [8]

    PathAgentBench: Benchmarking Evidence-Seeking Vision-Language Models on Whole-Slide Pathology Image

    URL https://arxiv.org/abs/2607.19261. Haotian Liu, Chunyuan Li, Qingyang Wu, and Yong Jae Lee. Visual instruction tuning.Advances in neural information processing systems, 36:34892–34916,

    work page Pith review arXiv

  3. [10]

    12 Chi Phan, Tianyi Zhang, Qiaochu Xue, Yufeng Wu, Dan Hu, Zeyu Liu, Sudong Wang, and Yueming Jin

    Accessed: 2026-07-24. 12 Chi Phan, Tianyi Zhang, Qiaochu Xue, Yufeng Wu, Dan Hu, Zeyu Liu, Sudong Wang, and Yueming Jin. Enhancing pathological vlms with cross-scale reasoning,

    work page 2026

  4. [11]

    Enhancing Pathological VLMs with Cross-scale Reasoning

    URL https://arxiv.org/abs/ 2606.17412. Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. Learning transferable visual models from natural language supervision. InInternational conference on machine learning, pages 8748–8763. PmLR,

    work page Pith review arXiv

  5. [12]

    Mehmet Saygin Seyfioglu, Wisdom O Ikezogwo, Fatemeh Ghezloo, Ranjay Krishna, and Linda Shapiro

    URL https://arxiv.org/abs/1707.06347. Mehmet Saygin Seyfioglu, Wisdom O Ikezogwo, Fatemeh Ghezloo, Ranjay Krishna, and Linda Shapiro. Quilt-llava: Visual instruction tuning by extracting localized narratives from open-source histopathology videos. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 13183–13192,

    arXiv

  6. [13]

    Pathmmu: A massive multimodal expert-level benchmark for understanding and reasoning in pathology

    Yuxuan Sun, Hao Wu, Chenglu Zhu, Sunyi Zheng, Qizi Chen, Kai Zhang, Yunlong Zhang, Dan Wan, Xiaoxiao Lan, Mengyue Zheng, et al. Pathmmu: A massive multimodal expert-level benchmark for understanding and reasoning in pathology. InEuropean Conference on Computer Vision, pages 56–73. Springer, 2024a. Yuxuan Sun, Chenglu Zhu, Sunyi Zheng, Kai Zhang, Lin Sun, ...

    work page 2025

  7. [14]

    URL https://doi.org/10.1038/s41551-026-01739-y

    doi: 10.1038/s41551-026-01739-y. URL https://doi.org/10.1038/s41551-026-01739-y. Yulin Wang, Yang Yue, Yang Yue, Huanqian Wang, Haojun Jiang, Yizeng Han, Zanlin Ni, Yifan Pu, Minglei Shi, Rui Lu, et al. Emulating human-like adaptive vision for efficient and flexible machine visual perception.Nature Machine Intelligence, pages 1–19,

    doi:10.1038/s41551-026-01739-y

  8. [15]

    An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al

    URL https://arxiv.org/abs/2506.20964. An Yang, Anfeng Li, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chang Gao, Chengen Huang, Chenxu Lv, et al. Qwen3 technical report,

    arXiv

Show all 17 references
  1. [16]

    13 Chunze Yang, Qidong Liu, Wenjie Zhao, Yue Tang, Jiusong Ge, Di Zhang, Jiashuai Liu, Lei Wu, Junbo Lu, Ni Zhang, Xian Wu, Zeyu Gao, and Chen Li

    URL https://arxiv.org/abs/ 2505.09388. 13 Chunze Yang, Qidong Liu, Wenjie Zhao, Yue Tang, Jiusong Ge, Di Zhang, Jiashuai Liu, Lei Wu, Junbo Lu, Ni Zhang, Xian Wu, Zeyu Gao, and Chen Li. Pathnavigate: A training-free pathology agent with surprise-guided scan and shared slide me...

    arXiv

  2. [17]

    Wenchuan Zhang, Penghao Zhang, Jingru Guo, Tao Cheng, Jie Chen, Shuwan Zhang, Zhang Zhang, Yuhao Yi, and Hong Bu

    URL https://arxiv.org/abs/2605.23559. Wenchuan Zhang, Penghao Zhang, Jingru Guo, Tao Cheng, Jie Chen, Shuwan Zhang, Zhang Zhang, Yuhao Yi, and Hong Bu. Patho-r1: A multimodal reinforcement learning-based pathology expert reasoner. InProceedings of the AAAI Conference on Artifi...

    arXiv

  3. [2017]

    Chunyuan Li, Cliff Wong, Sheng Zhang, Naoto Usuyama, Haotian Liu, Jianwei Yang, Tristan Naumann, Hoifung Poon, and Jianfeng Gao

    URL https://arxiv.org/abs/1412.6980. Chunyuan Li, Cliff Wong, Sheng Zhang, Naoto Usuyama, Haotian Liu, Jianwei Yang, Tristan Naumann, Hoifung Poon, and Jianfeng Gao. Llava-med: Training a large language-and-vision assistant for biomedicine in one day.Advances in Neural Informa...

    arXiv

  4. [2019]

    Ming Y Lu, Drew FK Williamson, Tiffany Y Chen, Richard J Chen, Matteo Barbieri, and Faisal Mahmood

    URL https: //arxiv.org/abs/1711.05101. Ming Y Lu, Drew FK Williamson, Tiffany Y Chen, Richard J Chen, Matteo Barbieri, and Faisal Mahmood. Data-efficient and weakly supervised computational pathology on whole-slide images. Nature Biomedical Engineering, 5(6):555–570,

    arXiv

  5. [2020]

    Chenshen Huang, Haoyun Xia, Xi Xiao, Hong Chen, Yiqing Jiang, Yahui Lyu, Zhizhan Ni, Tianyang Wang, Ning Wang, and Qi Huang

    URL https://arxiv.org/abs/2003.10286. Chenshen Huang, Haoyun Xia, Xi Xiao, Hong Chen, Yiqing Jiang, Yahui Lyu, Zhizhan Ni, Tianyang Wang, Ning Wang, and Qi Huang. Geometric multi-instance learning for weakly supervised gastric cancer segmentation.npj Digital Medicine,

    2003 arXiv

  6. [2023]

    Shuai Bai, Yuxuan Cai, Ruizhe Chen, Keqin Chen, Xionghui Chen, Zesen Cheng, Lianghao Deng, Wei Ding, Chang Gao, Chunjiang Ge, et al

    URL https://arxiv.org/abs/2308.12966. Shuai Bai, Yuxuan Cai, Ruizhe Chen, Keqin Chen, Xionghui Chen, Zesen Cheng, Lianghao Deng, Wei Ding, Chang Gao, Chunjiang Ge, et al. Qwen3-vl technical report,

    arXiv

  7. [2024]

    Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, Ruoyu Zhang, Shirong Ma, Xiao Bi, et al

    URL https://arxiv.org/abs/2407.21783. Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Peiyi Wang, Qihao Zhu, Runxin Xu, Ruoyu Zhang, Shirong Ma, Xiao Bi, et al. Deepseek-r1 incentivizes reasoning in llms through reinforce- ment learning.Nature, 645(8081):633–638,

    arXiv

  8. [2025]

    Thomas A

    URL https: //arxiv.org/abs/2511.21631. Thomas A. Buckley, Kian R. Weihrauch, Katherine Latham, Andrew Z. Zhou, Padmini A. Manrai, and Arjun K. Manrai. Navigating gigapixel pathology images with large multimodal models,

    arXiv

  9. [2026]

    Gabriele Campanella, Matthew G Hanna, Luke Geneslaw, Allen Miraflor, Vitor Werneck Krauss Silva, Klaus J Busam, Edi Brogi, Victor E Reuter, David S Klimstra, and Thomas J Fuchs

    URL https://arxiv.org/abs/2511.19652. Gabriele Campanella, Matthew G Hanna, Luke Geneslaw, Allen Miraflor, Vitor Werneck Krauss Silva, Klaus J Busam, Edi Brogi, Victor E Reuter, David S Klimstra, and Thomas J Fuchs. Clinical-grade computational pathology using weakly supervise...

    arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.