pith. machine review for the scientific record. sign in

arxiv: 2604.22026 · v2 · submitted 2026-04-23 · 💻 cs.AI · cs.CY· cs.DL

Recognition: no theorem link

Rethinking Publication: A Certification Framework for AI-Enabled Research

Yang Lu , Rabimba Karanjai , Lei Xu , Weidong Shi
Authors on Pith no claims yet

Pith reviewed 2026-05-13 07:56 UTC · model grok-4.3

classification 💻 cs.AI cs.CYcs.DL
keywords AI-generated researchpublication frameworkhuman contribution categoriesknowledge certificationautomated researchpeer reviewdecoupling authorshipbenchmark slots
0
0 comments X

The pith

A two-layer framework certifies AI-generated research by separating knowledge validity from human contribution.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper argues that the traditional publication system assumes all research comes from human authors, but AI pipelines can now produce work that meets standards for quality, novelty, and rigor. It proposes a certification approach with one layer that checks whether the knowledge claim is sound and a second layer that classifies the level of human involvement. This separation matters because it lets existing journals and conferences handle automated or partially automated submissions without creating new institutions. The framework responds to the fact that valid knowledge can now be generated even when human authorship is partial or uncertain. Dedicated benchmark slots for fully disclosed automated research would give a clear publication route and help reviewers refine their assessments over time.

Core claim

The publication system has historically certified both the validity of knowledge claims and the fact that they were produced by humans. AI research pipelines now allow the generation of valid knowledge with varying degrees of human involvement. The proposed two-layer framework first evaluates the soundness of the knowledge claim and then assesses the level of human contribution using three categories: Category A for work reachable by automated pipelines, Category B for work requiring human direction at identifiable stages, and Category C for work that goes beyond current pipeline capability especially in problem formulation. Dedicated benchmark slots for fully disclosed automated research of

What carries the argument

The two-layer certification framework that evaluates knowledge soundness in one layer and classifies the level of human contribution into three categories in the other layer.

If this is right

  • Journals and conferences can evaluate AI-enabled submissions using their current editorial systems.
  • Attribution uncertainty does not block assessment of the knowledge claim itself.
  • Human contributions that push beyond current AI capabilities, especially in problem formulation, receive recognition based on epistemic value.
  • Benchmark slots for fully automated research create transparent publication routes and supply data for refining category judgments over time.
  • Reviewers gain a structured method for making consistent judgments as AI capabilities change.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This separation could shift research incentives toward using AI for routine tasks while focusing human effort on novel problem formulation.
  • Data from the benchmark slots could empirically map where AI pipelines succeed or fail in research production.
  • The model might extend to other non-human contributors, such as large automated collaborations, in open science practices.
  • Adoption would likely require reviewer training to achieve reliable and consistent category identification across submissions.

Load-bearing premise

Reviewers can consistently and reliably identify the three categories of human contribution in practice, and journals will adopt the framework without needing new institutions or major process changes.

What would settle it

A test in which multiple independent reviewers apply the framework to the same set of AI-generated papers and show low agreement on category assignments, or journals refuse to integrate it because they require substantial new processes.

read the original abstract

AI research pipelines can now generate academic work that may satisfy existing peer review standards for quality, novelty, and methodological rigor. However, the publication system was built around the assumption that research is produced by human authors. It therefore lacks a clear way to evaluate work when the knowledge claim may be valid but the producer is partly or fully automated. This paper proposes a two-layer certification framework for AI-generated research. The first layer evaluates whether the knowledge claim is sound. The second layer evaluates the level of human contribution. This separation allows journals and conferences to assess pipeline-generated work more consistently without creating new institutions. The framework uses normative analysis, conceptual design, and dry-run validation against representative submission cases. It classifies human contribution into three categories: Category A, where the work is reachable by an automated pipeline; Category B, where human direction is required at identifiable stages; and Category C, where the work goes beyond current pipeline capability, especially at the problem-formulation stage. The paper also proposes dedicated benchmark slots for fully disclosed automated research. These slots would provide a transparent publication path and help reviewers calibrate judgments over time. The key argument is that publication has historically certified two things at once: that the knowledge is valid and that a human produced it. AI research pipelines separate these two claims. By decoupling knowledge certification from authorship attribution, the proposed framework responds to a structural change already underway. It can be implemented within existing editorial systems, works even when attribution is uncertain, and recognizes human frontier contribution based on epistemic value rather than human origin alone.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes a two-layer certification framework for AI-enabled research. The first layer evaluates whether a knowledge claim is sound according to existing peer-review standards. The second layer classifies the level of human contribution into three categories: Category A (work reachable by current automated pipelines), Category B (human direction required at identifiable stages), and Category C (work exceeds current pipeline capabilities, especially at problem formulation). The paper argues that publication has historically certified both validity and human authorship simultaneously, and that decoupling these allows consistent evaluation of pipeline-generated work within existing editorial systems. It also proposes dedicated benchmark slots for fully disclosed automated submissions to aid calibration.

Significance. If operationalized, the framework would address a structural shift in research production by allowing journals to certify epistemic value independently of authorship attribution. The normative separation of layers and the suggestion of benchmark slots provide a concrete path for handling uncertain attribution without new institutions. The conceptual design is grounded in analysis of current AI pipeline capabilities, though its significance depends on whether the categories can be applied reliably in practice.

major comments (2)
  1. [Framework proposal and category definitions] The section defining the three contribution categories (described in the abstract and framework proposal): the categories are introduced via normative analysis and high-level dry-run validation against representative cases, but no explicit decision criteria, edge-case rules, or annotated examples are supplied. This directly undermines the central claim that the framework 'can be implemented within existing editorial systems' and 'works even when attribution is uncertain,' because reviewers lack reproducible instructions for assigning A/B/C labels.
  2. [Validation and implementation discussion] The paragraph on dry-run validation (mentioned in the abstract and methods description): only a high-level reference to validation against representative submission cases is provided, with no reported inter-rater agreement metrics, disagreement cases, or quantitative outcomes. Without such data, the weakest assumption—that reviewers can reliably distinguish the categories—remains untested, making the decoupling of knowledge certification from authorship attribution non-operational as stated.
minor comments (1)
  1. [Abstract and Category C definition] The abstract states that the framework 'recognizes human frontier contribution based on epistemic value rather than human origin alone,' but the manuscript does not clarify how Category C contributions would be distinguished from Category B in cases of partial automation at the problem-formulation stage.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive comments and positive assessment of the framework's potential. We address each major comment point by point below, indicating revisions where we agree changes strengthen the manuscript.

read point-by-point responses

  1. Referee: The section defining the three contribution categories (described in the abstract and framework proposal): the categories are introduced via normative analysis and high-level dry-run validation against representative cases, but no explicit decision criteria, edge-case rules, or annotated examples are supplied. This directly undermines the central claim that the framework 'can be implemented within existing editorial systems' and 'works even when attribution is uncertain,' because reviewers lack reproducible instructions for assigning A/B/C labels.

    Authors: We agree that the current presentation of the categories would benefit from greater operational detail to support reliable implementation. In the revised manuscript we will add an explicit subsection containing decision criteria (including a decision tree for borderline cases), rules for handling uncertain attribution, and three to four annotated examples per category drawn from representative submission types. These additions will supply the reproducible instructions needed for reviewers while preserving the normative grounding of the framework. revision: yes

  2. Referee: The paragraph on dry-run validation (mentioned in the abstract and methods description): only a high-level reference to validation against representative submission cases is provided, with no reported inter-rater agreement metrics, disagreement cases, or quantitative outcomes. Without such data, the weakest assumption—that reviewers can reliably distinguish the categories—remains untested, making the decoupling of knowledge certification from authorship attribution non-operational as stated.

    Authors: The dry-run validation was conducted internally as a conceptual illustration against representative cases rather than a formal empirical study. We acknowledge that quantitative inter-rater metrics would provide stronger evidence of reliability. In revision we will expand the validation paragraph to describe the specific cases examined, note the main points of internal disagreement that arose, and explain how they were resolved. However, a controlled inter-rater study lies outside the scope of this conceptual proposal; we will add a brief note that such empirical calibration is a natural next step enabled by the benchmark slots we propose. revision: partial

standing simulated objections not resolved
  • Quantitative inter-rater agreement metrics from a formal empirical study, as the manuscript is a conceptual framework proposal and does not contain primary data collection or controlled testing.

Circularity Check

0 steps flagged

Normative framework proposal shows no circularity in derivation

full rationale

The paper advances a two-layer certification framework via normative analysis and conceptual design. Categories A/B/C are introduced as definitional distinctions based on the level of human contribution required, without any equations, fitted parameters, or predictions that reduce to inputs by construction. No self-citations appear as load-bearing premises, and the dry-run validation is presented as illustrative rather than a derivation step. The central decoupling argument follows directly from the stated problem analysis without self-referential loops or imported uniqueness theorems.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The framework depends on the assumption that knowledge claims can be evaluated independently of production method and introduces new conceptual categories without external empirical grounding.

axioms (2)
  • domain assumption Existing peer review standards can evaluate the soundness of a knowledge claim independently of authorship or production method.
    Invoked to justify the first layer of the framework.
  • ad hoc to paper Human contribution can be reliably classified into three discrete categories based on current automated pipeline capabilities.
    The A/B/C categories are defined in the paper to operationalize the second layer.
invented entities (1)
  • Category A, B, and C human contribution levels no independent evidence
    purpose: To classify the extent of human involvement for certification purposes.
    New conceptual categories introduced by the framework.

pith-pipeline@v0.9.0 · 5583 in / 1411 out tokens · 42387 ms · 2026-05-13T07:56:56.214171+00:00 · methodology

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

50 extracted references · 50 canonical work pages · 2 internal anchors

  1. [1]

    Introducing fars: A fully automated research system, 2026

    Analemma Intelligence . Introducing fars: A fully automated research system, 2026. URL https://analemma.ai/blog/introducing-fars/

    work page 2026

  2. [2]

    Fonseca, Robert Petryszak, Irene Papatheodorou, Ugis Sarkans, and Alvis Brazma

    Awais Athar, Anja F \"u llgrabe, Nancy George, Haider Iqbal, Laura Huerta, Ahmed Ali, Claire Snow, Nuno A. Fonseca, Robert Petryszak, Irene Papatheodorou, Ugis Sarkans, and Alvis Brazma. ArrayExpress update---from bulk to single-cell expression data. Nucleic Acids Research, 47 0 (D1): 0 D711--D715, 2019. doi:10.1093/nar/gky964

    work page doi:10.1093/nar/gky964 2019

  3. [3]

    Authorship and AI tools

    COPE Council . Authorship and AI tools. Committee on Publication Ethics, 2023. URL https://doi.org/10.24318/cCVRZBms

    work page doi:10.24318/ccvrzbms 2023

  4. [4]

    Ron Edgar, Michael Domrachev, and Alex E. Lash. Gene expression omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Research, 30 0 (1): 0 207--210, 2002. doi:10.1093/nar/30.1.207

    work page doi:10.1093/nar/30.1.207 2002

  5. [5]

    Ehrmann, A

    M. Ehrmann, A. Hamdi, E. Linhares Pontes, M. Romanello, and A. Doucet. Named entity recognition and classification in historical documents: A survey. ACM Computing Surveys, 56 0 (2): 0 Article 27, 2023. doi:10.1145/3604931. URL https://doi.org/10.1145/3604931

    work page doi:10.1145/3604931 2023

  6. [6]

    A Theory of Cognitive Dissonance

    Leon Festinger. A Theory of Cognitive Dissonance. Stanford University Press, Stanford, CA, 1957

    work page 1957

  7. [7]

    Carlsmith

    Leon Festinger and James M. Carlsmith. Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58 0 (2): 0 203--210, 1959. doi:10.1037/h0041593

    work page doi:10.1037/h0041593 1959

  8. [8]

    Funk and Jason Owen-Smith

    Russell J. Funk and Jason Owen-Smith. A dynamic network measure of technological change. Management Science, 63 0 (3): 0 791--817, 2017. URL https://doi.org/10.1287/mnsc.2015.2366

    work page doi:10.1287/mnsc.2015.2366 2017

  9. [9]

    Scientific Authorship: Credit and Intellectual Property in Science

    Peter Galison and Mario Biagioli, editors. Scientific Authorship: Credit and Intellectual Property in Science. Routledge, New York, 2003

    work page 2003

  10. [10]

    Empowering biomedical discovery with ai agents

    Shanghua Gao, Ada Fang, Yepeng Huang, Valentina Giunchiglia, Ayush Noori, Jonathan Richard Schwarz, Yasha Ektefaie, Jovana Kondic, and Marinka Zitnik. Empowering biomedical discovery with ai agents. Cell, 187 22: 0 6125--6151, 2024. URL https://api.semanticscholar.org/CorpusID:268875818

    work page 2024

  11. [11]

    Alvin I. Goldman. Knowledge in a Social World. Oxford University Press, Oxford, 1999. ISBN 9780198238201

    work page 1999

  12. [12]

    Charles A. E. Goodhart. Monetary Theory and Practice: The UK Experience . Macmillan, 1984

    work page 1984

  13. [13]

    How close is ChatGPT to human experts? Comparison corpus, evaluation, and detection

    Biyang Guo, Xin Zhang, Ziyuan Wang, Minqi Jiang, Jinran Nie, Yuxuan Ding, Jianwei Yue, and Yupeng Wu. How close is ChatGPT to human experts? Comparison corpus, evaluation, and detection. arXiv:2301.07597, 2023

    work page arXiv 2023

  14. [14]

    Weinberg

    Douglas Hanahan and Robert A. Weinberg. Hallmarks of cancer: the next generation. Cell, 144 0 (5): 0 646--674, 2011. doi:10.1016/j.cell.2011.02.013

    work page doi:10.1016/j.cell.2011.02.013 2011

  15. [15]

    Academic journals' AI policies fail to curb the surge in AI -assisted academic writing

    Yongyuan He and Yi Bu. Academic journals' AI policies fail to curb the surge in AI -assisted academic writing. Proceedings of the National Academy of Sciences, 122 0 (10): 0 e2526734123, 2025. doi:10.1073/pnas.2526734123

    work page doi:10.1073/pnas.2526734123 2025

  16. [16]

    Heine, and Ara Norenzayan

    Joseph Henrich, Steven J. Heine, and Ara Norenzayan. The weirdest people in the world? Behavioral and Brain Sciences, 33 0 (2--3): 0 61--83, 2010. doi:10.1017/S0140525X0999152X

    work page doi:10.1017/s0140525x0999152x 2010

  17. [17]

    Neural Computation 9, 1735–1780

    Sepp Hochreiter and J \"u rgen Schmidhuber. Long short-term memory. Neural Computation, 9 0 (8): 0 1735--1780, 1997. doi:10.1162/neco.1997.9.8.1735

    work page doi:10.1162/neco.1997.9.8.1735 1997

  18. [18]

    The Man Who Loved Only Numbers: The Story of Paul Erd o s and the Search for Mathematical Truth

    Paul Hoffman. The Man Who Loved Only Numbers: The Story of Paul Erd o s and the Search for Mathematical Truth . Hyperion, 1998

    work page 1998

  19. [19]

    Holcombe

    Alex O. Holcombe. Contributorship, not authorship: Use CRediT to indicate who did what. Publications, 7 0 (3): 0 48, 2019

    work page 2019

  20. [20]

    Rasmussen, and David B

    Mohammad Hosseini, Lisa M. Rasmussen, and David B. Resnik. Using AI to write scholarly publications. Accountability in Research, 30 0 (6): 0 394--406, 2023

    work page 2023

  21. [21]

    The role of ChatGPT in scientific communication: Writing better scientific review articles

    Jiawei Huang and Ming Tan. The role of ChatGPT in scientific communication: Writing better scientific review articles. American Journal of Cancer Research, 13 0 (4): 0 1148--1154, 2023

    work page 2023

  22. [22]

    Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals

    ICMJE . Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals. International Committee of Medical Journal Editors, 2023. URL http://www.icmje.org/recommendations/

    work page 2023

  23. [23]

    The state and fate of linguistic diversity and inclusion in the NLP research community

    Pratik Joshi, Sebastin Santy, Amar Budhiraja, Kalika Bali, and Monojit Choudhury. The state and fate of linguistic diversity and inclusion in the NLP research community. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 6282--6293, 2020

    work page 2020

  24. [24]

    Highly accurate protein structure prediction with AlphaFold

    John Jumper, Richard Evans, Alexander Pritzel, Tim Green, Michael Figurnov, Olaf Ronneberger, et al. Highly accurate protein structure prediction with AlphaFold . Nature, 596: 0 583--589, 2021

    work page 2021

  25. [25]

    Science, Truth, and Democracy

    Philip Kitcher. Science, Truth, and Democracy. Oxford Studies in Philosophy of Science. Oxford University Press, New York, 2001. ISBN 9780195145830. doi:10.1093/0195145836.001.0001

    work page doi:10.1093/0195145836.001.0001 2001

  26. [26]

    Krendl and Bernice A

    Anne C. Krendl and Bernice A. Pescosolido. Countries and cultural differences in the stigma of mental illness: the east--west divide. Journal of Cross-Cultural Psychology, 51 0 (2): 0 149--167, 2020. doi:10.1177/0022022119901297

    work page doi:10.1177/0022022119901297 2020

  27. [27]

    Monitoring AI -modified content at scale: A case study on the impact of ChatGPT on AI conference peer reviews

    Weixin Liang, Zachary Izzo, Yaohui Zhang, Haley Lepp, Hancheng Cao, Xuandong Zhao, others, and James Zou. Monitoring AI -modified content at scale: A case study on the impact of ChatGPT on AI conference peer reviews. arXiv:2403.07183. Published in Proceedings of ICML 2024, PMLR 235:29575--29620, 2024

    work page arXiv 2024

  28. [28]

    Helen E. Longino. The Fate of Knowledge. Princeton University Press, Princeton, NJ, 2002. ISBN 9780691088761

    work page 2002

  29. [29]

    The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery

    Chris Lu, Cong Lu, Robert Tjarko Lange, Jakob Foerster, Jeff Clune, and David Ha. The ai scientist: Towards fully automated open-ended scientific discovery, 2024. URL https://arxiv.org/abs/2408.06292

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  30. [30]

    Robert K. Merton. The normative structure of science. In Norman W. Storer, editor, The Sociology of Science: Theoretical and Empirical Investigations, pages 267--278. University of Chicago Press, 1973. Original work published 1942

    work page 1973

  31. [31]

    Nagata, Christopher D

    Jason M. Nagata, Christopher D. Otmar, Joan Shim, Priyadharshini Balasubramanian, Chloe M. Cheng, Elizabeth J. Li, Abubakr A. A. Al-Shoaibi, Iris Y. Shao, Kyle T. Ganson, Alexander Testa, Orsolya Kiss, Jinbo He, and Fiona C. Baker. Social media use and depressive symptoms during early adolescence. JAMA Network Open, 8 0 (5): 0 e2511704--e2511704, 05 2025....

    work page doi:10.1001/jamanetworkopen.2025.11704 2025

  32. [32]

    William M. Nauseef. Human neutrophils murine neutrophils: does it matter? Immunological Reviews, 314 0 (1): 0 442--456, 2023. doi:10.1111/imr.13154

    work page doi:10.1111/imr.13154 2023

  33. [33]

    Charles S. Peirce. Abduction and induction. In Charles Hartshorne and Paul Weiss, editors, Collected Papers of Charles Sanders Peirce, volume 5. Harvard University Press, 1935

    work page 1935

  34. [34]

    A Theory of Justice

    John Rawls. A Theory of Justice. Harvard University Press, 1971

    work page 1971

  35. [35]

    Foster, Ian T

    Andrey Rzhetsky, Jacob G. Foster, Ian T. Foster, and James A. Evans. Choosing experiments to accelerate collective discovery. Proceedings of the National Academy of Sciences, 112 0 (47): 0 14569--14574, 2015. doi:10.1073/pnas.1509757112. URL https://www.pnas.org/doi/abs/10.1073/pnas.1509757112

    work page doi:10.1073/pnas.1509757112 2015

  36. [36]

    System for systematic literature review using multiple ai agents: Concept and an empirical evaluation, 2025

    Abdul Malik Sami, Zeeshan Rasheed, Kai-Kristian Kemell, Muhammad Waseem, Terhi Kilamo, Mika Saari, Anh Nguyen Duc, Kari Systä, and Pekka Abrahamsson. System for systematic literature review using multiple ai agents: Concept and an empirical evaluation, 2025. URL https://arxiv.org/abs/2403.08399

    work page arXiv 2025

  37. [37]

    Shaw Warren, Alex G

    Junhee Seok, H. Shaw Warren, Alex G. Cuenca, Michael N. Mindrinos, Henry V. Baker, Weihong Xu, Daniel R. Richards, Grace P. McDonald-Smith, Hong Gao, Laura Hennessy, Celeste C. Finnerty, Cecilia M. L \'o pez, Shari Honari, Ernest E. Moore, Joseph P. Minei, Joseph Cuschieri, Paul E. Bankey, Jeffrey L. Johnson, Jason Sperry, Avery B. Nathens, Timothy R. Bil...

    work page doi:10.1073/pnas.1222878110 2013

  38. [38]

    Lederer, and Christophe Benoist

    Tal Shay, James A. Lederer, and Christophe Benoist. Genomic responses to inflammation in mouse models mimic humans: we concur, apples to oranges comparisons won't do. Proceedings of the National Academy of Sciences, 112 0 (4): 0 E346, 2015. doi:10.1073/pnas.1416629111

    work page doi:10.1073/pnas.1416629111 2015

  39. [39]

    What ChatGPT and generative AI mean for science

    Chris Stokel-Walker and Richard Van Noorden. What ChatGPT and generative AI mean for science. Nature, 614 0 (7947): 0 214--216, 2023

    work page 2023

  40. [40]

    Nature624(7990), 86–91 (2023) https://doi.org/10.1038/s41586-023-06734-w

    Nathan J. Szymanski, Bernardus Rendy, Yuxing Fei, Rishi E. Kumar, Tanjin He, David Milsted, Matthew J. McDermott, Max Gallant, Ekin Dogus Cubuk, Amil Merchant, Haegyeom Kim, Anubhav Jain, Christopher J. Bartel, Kristin Persson, Yan Zeng, and Gerbrand Ceder. An autonomous laboratory for the accelerated synthesis of inorganic materials. Nature, 624 0 (7990)...

    work page doi:10.1038/s41586-023-06734-w 2023

  41. [41]

    When combinations of humans and AI are useful: A systematic review and meta-analysis

    Michelle Vaccaro, Abdullah Almaatouq, and Thomas Malone. When combinations of humans and AI are useful: A systematic review and meta-analysis. Nature Human Behaviour, 8: 0 2293--2303, 2024. URL https://doi.org/10.1038/s41562-024-02024-1

    work page doi:10.1038/s41562-024-02024-1 2024

  42. [42]

    Gomez, ukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, ukasz Kaiser, and Illia Polosukhin. Attention is all you need. In Advances in Neural Information Processing Systems (NeurIPS), volume 30, pages 5998--6008, 2017

    work page 2017

  43. [43]

    Volkow, George F

    Nora D. Volkow, George F. Koob, Robert T. Croyle, Diana W. Bianchi, Joshua A. Gordon, Walter J. Koroshetz, Eliseo J. P\' e rez-Stable, Wilson T. Riley, Michelle H. Bloch, Kevin Conway, Barbara G. Deeds, Gayathri J. Dowling, Steven Grant, Kathleen D. Howlett, John A. Matochik, Gaillard W. Morgan, Margaret M. Murray, Antonio Noronha, Catherine Y. Spong, and...

    work page doi:10.1016/j.dcn.2017.10.002 2017

  44. [44]

    On the origin of cancer cells

    Otto Warburg. On the origin of cancer cells. Science, 123 0 (3191): 0 309--314, 1956. doi:10.1126/science.123.3191.309

    work page doi:10.1126/science.123.3191.309 1956

  45. [45]

    Testing of detection tools for AI -generated text

    Debora Weber-Wulff, Alla Anohina-Naumeca, Sonja Bjelobaba, Tom \'a s Folt \'y nek, Jean Guerrero-Dib, Olumide Popoola, others, and Lorna Waddington. Testing of detection tools for AI -generated text. International Journal of Educational Integrity, 19 0 (1): 0 26, 2023

    work page 2023

  46. [46]

    Lingfei Wu, Dashun Wang, and James A. Evans. Large teams develop and small teams disrupt science and technology. Nature, 566 0 (7744): 0 378--382, 2019

    work page 2019

  47. [47]

    Jones, and Brian Uzzi

    Stefan Wuchty, Benjamin F. Jones, and Brian Uzzi. The increasing dominance of teams in production of knowledge. Science, 316 0 (5827): 0 1036--1039, 2007

    work page 2007

  48. [48]

    The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search

    Yutaro Yamada, Robert Tjarko Lange, Cong Lu, Shengran Hu, Chris Lu, Jakob Foerster, Jeff Clune, and David Ha. The ai scientist-v2: Workshop-level automated scientific discovery via agentic tree search, 2025. URL https://arxiv.org/abs/2504.08066

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  49. [49]

    Virtues of the Mind: An Inquiry into the Nature of Virtue and the Ethical Foundations of Knowledge

    Linda Trinkaus Zagzebski. Virtues of the Mind: An Inquiry into the Nature of Virtue and the Ethical Foundations of Knowledge. Cambridge Studies in Philosophy. Cambridge University Press, Cambridge, 1996. ISBN 9780521570602. doi:10.1017/CBO9781139174763

    work page doi:10.1017/cbo9781139174763 1996

  50. [50]

    Zhavoronkov, D

    A. Zhavoronkov, D. Gennert, and J. Shi. From prompt to drug: Toward pharmaceutical superintelligence. ACS Central Science, 2026. doi:10.1021/acscentsci.5c01473. URL https://doi.org/10.1021/acscentsci.5c01473. Advance online publication

    work page doi:10.1021/acscentsci.5c01473 2026