Pith. sign in

REVIEW 4 major objections 4 minor 37 references

Self-control: A Better Conditional Mechanism for Masked Autoregressive Model

T0 review · 4 major / 4 minor · reviewed 2026-08-11 · deepseek-v4-flash

Pith's one-line read The paper proposes a self-control network that unifies text and image conditions with the generated image in a single self-attention sequence, replacing cross-attention for continuous autoregressive image generation.

desk verdict An incomplete draft: the proposed self-control mechanism is a small variation on existing concatenation conditioning, and the paper provides no experimental evidence for its central claim. read the letter →

arxiv 2412.13635 v1 pith:ODHQJE2O submitted 2024-12-18 cs.CV

classification cs.CV
keywords self-controlnetworkmaskedautoregressivemodelcontinuousimagegenerationconditionalself-attentionmultimodalfusiontext-to-imagevectorquantization
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 proposes replacing the cross-attention fusion stage in continuous autoregressive image generation with a single self-attention pass over a sequence that stacks text tokens, image-condition tokens, and the image tokens being generated. The claim is that this self-control network makes multimodal conditioning simpler and lets a continuous masked autoregressive model generate images that follow text and image conditions without the quality loss that vector-quantized codebooks introduce. A sympathetic reader would care because most autoregressive image models pay for discretization of the image lattice, and this proposal is a candidate mechanism for keeping autoregressive conditioning while generating in continuous space.

What carries the argument

The central object is the self-control network, a masked autoregressive transformer whose attention pattern is a hybrid of causal and bidirectional masks applied to a single concatenated sequence of text tokens, image condition tokens, and generated image tokens. The causal mask on text and on the cross-modal path enforces the autoregressive order required for coherent generation, while the bidirectional masks on image tokens let spatial context be read from both directions. This mixed-mask self-attention is what carries the argument: it is the single mechanism that replaces cross-attention-based conditional fusion.

What would settle it

Train the same continuous masked autoregressive backbone twice, once with the proposed mixed-mask self-control attention and once with a standard cross-attention conditional mechanism at equal parameter count, and compare FID and text-image alignment on a benchmark such as VEDAI or COCO; if self-control does not match or beat cross-attention on both metrics, the central claim fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes a conditional mechanism: concatenate condition tokens from text and image encoders with the tokens of the image to be generated, feed the whole sequence into a masked autoregressive transformer, and apply separate attention masks inside one self-attention computation. Text condition tokens are processed causally to respect temporal order, image condition tokens and generated image tokens are processed bidirectionally to capture spatial context, and cross-modal attention between the three groups is causal so that each generated token attends to all available conditions without allowing the conditions to reconstruct each other. The authors argue that this unified sequence removes the need for a dedicated cross-attention module and lets the model learn text-image fusion in the same representation space as generation, mitigating the reconstruction and quality penalties associated with vector quantization.

Load-bearing premise

The load-bearing premise is that feeding one self-attention pass a sequence of text tokens, image condition tokens, and generated image tokens will fuse the modalities at least as well as a dedicated cross-attention module; the paper motivates this premise but does not compare it against cross-attention.

Editorial extensions

If this is right

  • Continuous autoregressive text-to-image generation can be built without a separate cross-attention stack, so conditional fusion and generation share one transformer backbone.
  • Text and image conditions can be combined in the same autoregressive sequence, which makes multi-condition control such as text plus reference image a natural extension of the generation loop.
  • Avoiding vector-quantized codes removes the codebook reconstruction bottleneck that the paper identifies as the main quality limitation of discrete autoregressive models.
  • Because the masks are the only structural difference from an ordinary masked autoregressive transformer, the mechanism can be dropped into existing continuous autoregressive frameworks.

Reading between the lines

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

  • One test the paper leaves implicit is a controlled comparison of self-control attention against standard cross-attention at equal parameter and token counts; the included ablation varies masks within the self-control design rather than against a cross-attention baseline.
  • The same hybrid-mask sequence could plausibly extend to video or multi-view generation, where causal text prompts and bidirectional spatial tokens follow the same logic, though the paper does not discuss those settings.
  • Unifying all modalities in one self-attention pass increases the sequence length by the condition tokens, so the practical benefit depends on whether the removed cross-attention module saves more compute than the longer self-attention costs; the paper does not quantify this trade-off.
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 / 4 minor

Summary. The paper proposes a "self-control" conditional injection mechanism for continuous masked autoregressive image generation. Instead of using cross-attention to fuse text or image conditions with generated image tokens, it concatenates condition tokens and generated-image tokens into one sequence and applies a mixed attention mask: causal attention between condition text tokens, bidirectional attention among image condition tokens and among generated image tokens, and causal attention across modality groups. The authors argue that this unifies multimodal conditions and generated content in a single self-attention space, avoiding the quality loss from vector quantization and improving controllability relative to cross-attention. The manuscript contains a short literature review, an informal methodology section with schematic figures, and an experiments section that consists only of a table header with no numerical results.

Significance. If fully validated, the proposed mechanism would be a simple and potentially useful alternative to cross-attention for conditional image generation in continuous autoregressive models. The conceptual framing is clear, and the authors correctly identify vector-quantization-free continuous autoregressive models as an important recent direction. However, the paper currently delivers no empirical evidence whatsoever: no quantitative results, no baselines, no ablations, no training or inference protocol, and no code. The central claim is empirical, and as written it is untestable. The manuscript is closer to an extended abstract or a draft than to a complete research paper, and its contribution cannot be assessed at the standard expected for a journal.

major comments (4)
  1. [Section V, Table I] The entire experimental evaluation is missing. Section V contains only the subsection heading "Ablation experiments" and Table I with column headers for "text", "image", "multimodal", "FID", and "IS"; every numeric cell is empty. There are no results, no baselines, no evaluation protocol, and no comparison against cross-attention or against existing continuous autoregressive models such as MAR or Fluid. Since the paper's central claim is empirical ('better conditional mechanism'), this omission leaves the claim entirely unsupported.
  2. [Section IV.C and Eq. (1)] The proposed attention masks conflict with the autoregressive factorization stated in Eq. (1). Eq. (1) defines the joint distribution as a product of conditional probabilities p(x_i | x_1, ..., x_{i-1}), which requires that token i cannot attend to tokens j > i. Section IV.C, however, applies bidirectional attention to the generated image tokens (the orange lines in Fig. 3), allowing each generated token to attend to both earlier and later generated tokens. This is inconsistent with Eq. (1) unless a different inference procedure, such as parallel decoding with repeated visible-token replacement, is specified. The manuscript never specifies the training mask, the sampling loop, or the loss, so the mechanism is undefined and the claim cannot be verified or falsified.
  3. [Section IV.B] The decision to apply causal attention to text condition tokens is not justified and appears counterproductive. The full text prompt is known before image generation begins, so there is no sequential reason to prevent text tokens from attending to later text tokens; standard text-to-image autoregressive models encode the full prompt bidirectionally or with cross-attention. The assertion that causal attention 'more accurately understand[s] the textual description' is made without evidence, and no ablation comparing causal versus bidirectional text masking is provided. At minimum, the authors need to justify this design choice with an experiment or a formal argument.
  4. [Section IV.A and paper title] The paper's comparative claim against cross-attention is never tested. The title and abstract promise a mechanism that is 'better' than 'the conventional cross-attention-based conditional fusion mechanism', but nowhere in the paper is a cross-attention baseline defined or compared. Table I only varies the three mask options within the proposed self-attention framework. Without a controlled comparison against a cross-attention-based continuous autoregressive model under the same backbone, data, and training budget, the central contribution is unsubstantiated.
minor comments (4)
  1. [Index Terms] The phrase "Index Terms —Autoresressive" contains a typo; it should read "Autoregressive".
  2. [Section IV.A] The description of the autoregressive module as "MAE-style encoder-decoder" is too vague. The paper should specify architectural dimensions, patch size, token sequence length, diffusion loss details, and training hyperparameters, or cite the exact configuration used in MAR [6].
  3. [Section III.B, Eq. (2)] The notation in Eq. (2) is under-specified: the range of the product over k is not defined, and the sets X^k are not precisely characterized. The paper should define K and the partition formally.
  4. [Fig. 4] Figure 4, "Illustration of the visualization of attention", is never discussed in the text, and its relationship to the proposed method is unclear.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the self-control mechanism is under-specified and empirically unsupported, but nothing in the derivation reduces to its own inputs or to a load-bearing self-citation.

full rationale

The paper's central claim is that concatenating text condition tokens, image condition tokens, and generated image tokens into one sequence with mixed causal and bidirectional attention improves conditional control and image quality. That claim is not derived from any equation that equals its conclusion by construction, and no fitted parameter is renamed as a prediction. The method sections (IV A-D) merely describe an architecture choice, and the only experimental section (V A) contains Table I with an empty quantitative result, so the claim is untestable as written. This is a missing-evidence and under-specification problem, not a circularity problem. Citations to MAR [6], Fluid [7], and MAE [30] are external prior work used as architectural background, not self-citations that carry the argument, and no uniqueness theorem or prior author result is invoked to force the design. Because no step in the paper reduces by definition or by self-citation to its own input, the circularity score is 0.

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

The paper relies on standard autoregressive factorization and on the MAR/Fluid assumption that continuous tokens with diffusion loss are a valid substitute for vector quantization. Its own architectural claim, that mixed causal and bidirectional self-attention improves conditioning, is an unverified ad hoc assumption. No free parameters or invented entities are present because no experiments are reported.

assumptions (3)
  • standard math Autoregressive factorization p(x1,...,xn) = product_i p(x_i | x_<i) holds for image generation.
    Invoked in Section III-A Equation (1) and used as the basis for the sequence model.
  • domain assumption Continuous token representations with diffusion loss can replace vector quantization in autoregressive image models.
    Adopted from MAR and Fluid in Sections I and II; the paper does not re-derive or validate this.
  • ad hoc to paper Concatenating text and image condition tokens with generated tokens into a single self-attention sequence and using causal and bidirectional masks will improve conditional generation compared with cross-attention.
    The core architectural premise in Section IV; no experiments or theoretical argument support it.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Self-control: A Better Conditional Mechanism for Masked Autoregressive Model." pith.science (2026). https://pith.science/paper/ODHQJE2O

@misc{pith2026241213635,
  author       = {Pith},
  title        = {Pith review of: Self-control: A Better Conditional Mechanism for Masked Autoregressive Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ODHQJE2O}},
  note         = {Machine review of arXiv:2412.13635}
}
read the original abstract

Autoregressive conditional image generation algorithms are capable of generating photorealistic images that are consistent with given textual or image conditions, and have great potential for a wide range of applications. Nevertheless, the majority of popular autoregressive image generation methods rely heavily on vector quantization, and the inherent discrete characteristic of codebook presents a considerable challenge to achieving high-quality image generation. To address this limitation, this paper introduces a novel conditional introduction network for continuous masked autoregressive models. The proposed self-control network serves to mitigate the negative impact of vector quantization on the quality of the generated images, while simultaneously enhancing the conditional control during the generation process. In particular, the self-control network is constructed upon a continuous mask autoregressive generative model, which incorporates multimodal conditional information, including text and images, into a unified autoregressive sequence in a serial manner. Through a self-attention mechanism, the network is capable of generating images that are controllable based on specific conditions. The self-control network discards the conventional cross-attention-based conditional fusion mechanism and effectively unifies the conditional and generative information within the same space, thereby facilitating more seamless learning and fusion of multimodal features.

Figures

Figures reproduced from arXiv: 2412.13635 by the authors.

Figure 1
Figure 1. Illustration of the causal attention and bidirectional [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Illustration of self-control network [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Illustration of attention mechanisms in autoregressive module. [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Illustration of the visualization of attention. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

37 extracted references · 24 canonical work pages

  1. [1]

    Pixel recurrent neural networks,

    A. van den Oord, N. Kalchbrenner, and K. Kavukcuoglu, “Pixel recurrent neural networks,” in International Conference on Machine Learning ,

  2. [2]

    Conditional image generation with pixelcnn decoders,

    A. van den Oord, N. Kalchbrenner, L. Espeholt, K. Kavukcuoglu, O. Vinyals, and A. Graves, “Conditional image generation with pixelcnn decoders,” ArXiv, vol. abs/1606.05328, 2016. [Online]. Available: https://api.semanticscholar.org/CorpusID:14989939

  3. [3]

    Autoregressive model beats diffusion: Llama for scalable image generation,

    P. Sun, Y . Jiang, S. Chen, S. Zhang, B. Peng, P. Luo, and Z. Yuan, “Autoregressive model beats diffusion: Llama for scalable image generation,” ArXiv, vol. abs/2406.06525, 2024. [Online]. Available: https://api.semanticscholar.org/CorpusID:270371603

  4. [4]

    Magvit: Masked generative video transformer,

    L. Yu, Y . Cheng, K. Sohn, J. Lezama, H. Zhang, H. Chang, A. G. Hauptmann, M.-H. Yang, Y . Hao, I. Essa, and L. Jiang, “Magvit: Masked generative video transformer,” 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pp. 10 459–10 469,

  5. [5]

    Visual autoregressive modeling: Scalable image generation via next-scale prediction,

    K. Tian, Y . Jiang, Z. Yuan, B. Peng, and L. Wang, “Visual autoregressive modeling: Scalable image generation via next-scale prediction,” ArXiv, vol. abs/2404.02905, 2024. [Online]. Available: https://api.semanticscholar.org/CorpusID:268876071

  6. [6]

    Autoregressive image generation without vector quantization,

    T. Li, Y . Tian, H. Li, M. Deng, and K. He, “Autoregressive image generation without vector quantization,” ArXiv, vol. abs/2406.11838,

  7. [7]

    Fluid: Scaling autoregressive text-to-image generative models with continuous tokens,

    L. Fan, T. Li, S. Qin, Y . Li, C. Sun, M. Rubinstein, D. Sun, K. He, and Y . Tian, “Fluid: Scaling autoregressive text-to-image generative models with continuous tokens,” ArXiv, vol. abs/2410.13863, 2024. [Online]. Available: https://api.semanticscholar.org/CorpusID:273404213

  8. [8]

    Neural discrete representation learning,

    A. van den Oord, O. Vinyals, and K. Kavukcuoglu, “Neural discrete representation learning,” in Neural Information Processing Systems ,

Show all 37 references
  1. [9]

    Taming transformers for high-resolution image synthesis,

    P. Esser, R. Rombach, and B. Ommer, “Taming transformers for high-resolution image synthesis,” 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pp. 12 868–12 878,

  2. [10]

    Scalable autoregressive image generation with mamba,

    H. Li, J. Yang, K. Wang, X. Qiu, Y . Chou, X. Li, and G. Li, “Scalable autoregressive image generation with mamba,” ArXiv, vol. abs/2408.12245, 2024. [Online]. Available: https://api.semanticscholar. org/CorpusID:271924150

  3. [11]

    Maskgit: Masked generative image transformer,

    H. Chang, H. Zhang, L. Jiang, C. Liu, and W. T. Freeman, “Maskgit: Masked generative image transformer,” 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pp. 11 305–11 315,

  4. [12]

    Mage: Masked generative encoder to unify representation learning and image synthesis,

    T. Li, H. Chang, S. K. Mishra, H. Zhang, D. Katabi, and D. Krishnan, “Mage: Masked generative encoder to unify representation learning and image synthesis,” 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2142–2152, 2022. [Online]. Available: ht...

  5. [13]

    Givt: Generative infinite- vocabulary transformers,

    M. Tschannen, C. Eastwood, and F. Mentzer, “Givt: Generative infinite- vocabulary transformers,” ArXiv, vol. abs/2312.02116, 2023. [Online]. Available: https://api.semanticscholar.org/CorpusID:265610025

  6. [14]

    Generative adversarial nets,

    I. J. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. C. Courville, and Y . Bengio, “Generative adversarial nets,” in Neural Information Processing Systems , 2014. [Online]. Available: https://api.semanticscholar.org/CorpusID:261560300

  7. [15]

    Denoising diffusion probabilistic models,

    J. Ho, A. Jain, and P. Abbeel, “Denoising diffusion probabilistic models,” ArXiv, vol. abs/2006.11239, 2020. [Online]. Available: https://api.semanticscholar.org/CorpusID:219955663

  8. [16]

    Stackgan: Text to photo-realistic image synthesis with stacked generative adversarial networks,

    H. Zhang, T. Xu, H. Li, S. Zhang, X. Wang, X. Huang, and D. N. Metaxas, “Stackgan: Text to photo-realistic image synthesis with stacked generative adversarial networks,” 2017 IEEE International Conference on Computer Vision (ICCV), pp. 5908–5916, 2016. [Online]. Available: htt...

  9. [17]

    Available: https://api.semanticscholar.org/CorpusID: 246680316

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 246680316

  10. [18]

    Sd-gan: Semantic decomposition for face image synthesis with discrete attribute,

    K. Zhou, X. Zhu, D. Gao, K. Lee, X. Li, and X.-C. Yin, “Sd-gan: Semantic decomposition for face image synthesis with discrete attribute,” Proceedings of the 30th ACM International Conference on Multimedia ,

  11. [19]

    Glide: Towards photorealistic image generation and editing with text-guided diffusion models,

    A. Nichol, P. Dhariwal, A. Ramesh, P. Shyam, P. Mishkin, B. McGrew, I. Sutskever, and M. Chen, “Glide: Towards photorealistic image generation and editing with text-guided diffusion models,” in International Conference on Machine Learning , 2021. [Online]. Available: https://a...

  12. [20]

    Photorealistic text-to-image diffusion models with deep language understanding,

    C. Saharia, W. Chan, S. Saxena, L. Li, J. Whang, E. L. Denton, S. K. S. Ghasemipour, B. K. Ayan, S. S. Mahdavi, R. G. Lopes, T. Salimans, J. Ho, D. J. Fleet, and M. Norouzi, “Photorealistic text-to-image diffusion models with deep language understanding,” ArXiv, vol. abs/2205....

  13. [21]

    Hierarchical text-conditional image generation with clip latents,

    A. Ramesh, P. Dhariwal, A. Nichol, C. Chu, and M. Chen, “Hierarchical text-conditional image generation with clip latents,” ArXiv, vol. abs/2204.06125, 2022. [Online]. Available: https://api. semanticscholar.org/CorpusID:248097655

  14. [22]

    High-resolution image synthesis with latent diffusion models,

    R. Rombach, A. Blattmann, D. Lorenz, P. Esser, and B. Ommer, “High-resolution image synthesis with latent diffusion models,” 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pp. 10 674–10 685, 2021. [Online]. Available: https://api.semanticscholar.o...

  15. [23]

    Attngan: Fine-grained text to image generation with attentional generative adversarial networks,

    T. Xu, P. Zhang, Q. Huang, H. Zhang, Z. Gan, X. Huang, and X. He, “Attngan: Fine-grained text to image generation with attentional generative adversarial networks,” 2018 IEEE/CVF Conference on JOURNAL OF LATEX CLASS FILES, VOL. 18, NO. 9, SEPTEMBER 2020 5 TABLE I: Quantitative...

  16. [24]

    All are worth words: A vit backbone for diffusion models,

    F. Bao, S. Nie, K. Xue, Y . Cao, C. Li, H. Su, and J. Zhu, “All are worth words: A vit backbone for diffusion models,” 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pp. 22 669–22 679, 2022. [Online]. Available: https://api.semanticscholar.org/Cor...

  17. [25]

    Available: https://api.semanticscholar.org/CorpusID: 250533889

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 250533889

  18. [26]

    Dreambooth: Fine tuning text-to-image diffusion models for subject- driven generation,

    N. Ruiz, Y . Li, V . Jampani, Y . Pritch, M. Rubinstein, and K. Aberman, “Dreambooth: Fine tuning text-to-image diffusion models for subject- driven generation,” 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pp. 22 500–22 510, 2022. [Online]. Ava...

  19. [27]

    Adding conditional control to text- to-image diffusion models,

    L. Zhang, A. Rao, and M. Agrawala, “Adding conditional control to text- to-image diffusion models,” 2023 IEEE/CVF International Conference on Computer Vision (ICCV) , pp. 3813–3824, 2023. [Online]. Available: https://api.semanticscholar.org/CorpusID:256827727

  20. [28]

    Zero-shot text-to-image generation,

    A. Ramesh, M. Pavlov, G. Goh, S. Gray, C. V oss, A. Radford, M. Chen, and I. Sutskever, “Zero-shot text-to-image generation,” ArXiv, vol. abs/2102.12092, 2021. [Online]. Available: https://api.semanticscholar. org/CorpusID:232035663

  21. [29]

    Cogview: Mastering text-to-image generation via transformers,

    M. Ding, Z. Yang, W. Hong, W. Zheng, C. Zhou, D. Yin, J. Lin, X. Zou, Z. Shao, H. Yang, and J. Tang, “Cogview: Mastering text-to-image generation via transformers,” in Neural Information Processing Systems , 2021. [Online]. Available: https: //api.semanticscholar.org/CorpusID:...

  22. [30]

    Scalable diffusion models with transformers,

    W. S. Peebles and S. Xie, “Scalable diffusion models with transformers,” 2023 IEEE/CVF International Conference on Computer Vision (ICCV) , pp. 4172–4182, 2022. [Online]. Available: https: //api.semanticscholar.org/CorpusID:254854389

  23. [32]

    Lora: Low-rank adaptation of large language models,

    J. E. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, and W. Chen, “Lora: Low-rank adaptation of large language models,” ArXiv, vol. abs/2106.09685, 2021. [Online]. Available: https://api.semanticscholar.org/CorpusID:235458009

  24. [37]

    Masked autoencoders are scalable vision learners,

    K. He, X. Chen, S. Xie, Y . Li, P. Doll’ar, and R. B. Girshick, “Masked autoencoders are scalable vision learners,” 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) , pp. 15 979–15 988, 2021. [Online]. Available: https://api.semanticscholar. org/Corpu...

  25. [2016]

    Available: https://api.semanticscholar.org/CorpusID: 8142135

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 8142135

  26. [2017]

    Available: https://api.semanticscholar.org/CorpusID: 20282961

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 20282961

  27. [2020]

    Available: https://api.semanticscholar.org/CorpusID: 229297973

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 229297973

  28. [2022]

    Available: https://api.semanticscholar.org/CorpusID: 254563906

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 254563906

  29. [2024]

    Available: https://api.semanticscholar.org/CorpusID: 270560593

    [Online]. Available: https://api.semanticscholar.org/CorpusID: 270560593

Pith tools

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