Pith. sign in

REVIEW 3 major objections 3 minor 167 references

The Cake that is Intelligence and Who Gets to Bake it: An AI Analogy and its Implications for Participation

T0 review · 3 major / 3 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read Even if everyone agreed on what good AI should look like, the statistical foundations of machine learning would still make it extremely hard to build it.

desk verdict A genuinely useful reframing of the AI lifecycle through the cake analogy, but the central causal claim is asserted more strongly than the evidence supports. read the letter →

arxiv 2502.03038 v2 pith:YJN4X3JF submitted 2025-02-05 cs.AI cs.CYcs.LG

Martin Mundt , Anaelia Ovalle , Felix Friedrich , A Pranav , Subarnaduti Paul , Manuel Brack , Kristian Kersting , William Agnew This is my paper · ORCID
classification cs.AIcs.CYcs.LG
keywords AIlifecyclecakeanalogyi.i.d.assumptionhomogenizationcontinuallearningaveragedmetricssurrogateobjectivesparticipatory
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 social problems of AI are not just a matter of bad intentions or weak regulation; they are baked into the statistical and architectural assumptions of machine learning itself. The authors extend Yann LeCun's "intelligence is a cake" metaphor from a structural comparison of learning paradigms into a full lifecycle analogy covering data sourcing, recipe design, training, evaluation, and distribution. For each stage they identify a technical foundation -- the i.i.d. data assumption, homogenized model architectures, single-pass training, averaged evaluation metrics, and surrogate objectives -- and show how it constrains attempts to make AI more participatory and just. The central claim is that even if consensus on instrumental and normative aims existed, the present technical foundations would make it tremendously difficult to translate benevolent aims into practice. The paper closes with technical recommendations for each stage, aimed at opening the design process to broader participation.

What carries the argument

The central object is the re-conceptualized AI-cake analogy, a stage-by-stage mapping of the AI lifecycle (ingredients=data, recipe=model instructions, baking=training, tasting=evaluation, selling=distribution) onto statistical machine learning assumptions. The load-bearing technical mechanisms are: the i.i.d. factorization of the likelihood, which turns a joint distribution into a product and hides inter-sample dependencies; the cascaded composite hypothesis $h_{\theta}(x) = T_L(T_{L-1}(\cdots T_1(x)))$ representing homogenized deep models; iterative stochastic gradient updates $\theta \leftarrow \theta - \eta \nabla \mathcal{L}_{\tau}(\theta)$ that cannot incorporate new knowledge without catastrophic forgetting; averaged losses $\mathcal{L}(\theta) = \frac{1}{N}\sum_n \mathcal{L}_n(\theta)$ that obscure underrepresented groups; and surrogate objectives such as reconstruction error or next-token prediction that stand in for unmeasurable goals. These mechanisms carry the argument by showing that each social ramification identified in the analogy has a corresponding mathematical constraint that makes change difficult.

What would settle it

A concrete test would be to build or find an AI system that relaxes all four technical constraints -- explicit modeling of non-i.i.d. dependencies, a non-homogenized architecture with continual learning, individualized rather than averaged evaluation, and objectives directly tied to human-defined outcomes -- and then measure whether the social harms documented by the paper (e.g., stereotype collapse, misclassification of marginalized groups, overselling) disappear or substantially decrease.

Watch

Extended reading notes

Core claim

The paper's central discovery is that the cake analogy, taken literally as a lifecycle rather than a structural metaphor, exposes a tight coupling between social outcomes and technical choices. Sourcing ingredients maps to dataset acquisition and the i.i.d. assumption that hides non-i.i.d. correlations; recipes map to the homogenization of foundation models and software frameworks; baking maps to a single unidirectional training process in which post-hoc modification is only superficial; tasting maps to averaged evaluation that collapses diversity; and selling maps to the "abundant surrogate impasse," where optimization objectives, evaluation measures, and real-world goals diverge. The authors assert that these technical underpinnings actively impede benevolent aims, so participation and ethical reform must engage with machine learning foundations, not only with governance or representation.

Load-bearing premise

The paper's argument depends on the assumption that the cake-analogy mapping is faithful enough that its technical choices -- i.i.d. data, homogenized architectures, averaged losses, and surrogate objectives -- are genuinely root causes of the social harms it describes, rather than symptoms of market or political forces.

Editorial extensions

If this is right

  • If the i.i.d. assumption is computationally necessary, then dataset documentation and de-duplication are not optional extras but prerequisites for any reliable claim about what a model learned.
  • If post-training modification is largely superficial, then audits and red-teaming identify problems but cannot be the primary mechanism for fixing them; fundamental changes require retraining, with its costs.
  • If homogenized frameworks actively penalize alternative architectures, then calls for diverse model designs need simultaneous investment in new software frameworks.
  • If averaged metrics erase group-level harms, then reporting single benchmark numbers is itself a source of misleading reassurance about fairness and capability.
  • If the surrogate impasse is real, then claims about AI capability should be tied to outcomes, not to proxy losses, and evaluation needs causal or counterfactual assessment.

Reading between the lines

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

  • The paper's argument implies a testable hierarchy: interventions that change data collection and training objectives should have larger downstream social effects than post-hoc alignment or evaluation reform.
  • A natural extension is to quantify the "abundant surrogate impasse" by measuring divergence between proxy-optimized models' behavior and human-judged outcomes across domains; if this divergence is consistently large, surrogate objectives are a primary driver of overselling.
  • The analogy suggests a design principle: treat AI systems as artifacts that must be cheaply "re-baked," making continual learning and modular architectures preconditions for participatory change rather than mere efficiency gains.
  • The paper implicitly predicts that participatory data stewardship and evaluation dashboards will fail to change system behavior unless paired with changes in the i.i.d., averaging, and surrogate assumptions.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 3 minor

Summary. This position paper extends Yann LeCun's cake metaphor from a structural analogy among learning paradigms to the full AI lifecycle: ingredient sourcing (data), recipe design (model architectures and training procedures), baking (training), tasting (evaluation), and selling (deployment). The authors argue that social harms in AI are intertwined with foundational statistical and optimization assumptions—i.i.d. data, homogenized architectures, single-cycle training, averaged losses, and surrogate objectives—and that these technical assumptions make it 'tremendously challenging' to translate benevolent aims into practice even if normative consensus existed. The paper closes with technical recommendations mapped to each lifecycle stage. The mathematical content (Eqs. 1–7) is standard and used illustratively; the paper offers no new formal results or empirical evaluation.

Significance. If the central claim is accepted, the paper provides a valuable bridge between critical AI studies and the technical ML community, giving practitioners a concrete vocabulary for where participation and ethical intervention can attach to the modeling pipeline. Its strengths are breadth of cited literature, an explicit limitation section, and a clear set of actionable recommendations. The paper is best read as a conceptual mapping and research agenda rather than an empirical demonstration; the load-bearing causal inference—that technical assumptions are a major barrier to social improvement—needs either more support or a more cautious framing.

major comments (3)
  1. [Section 3, introductory paragraph] The central claim that 'the present technical foundations make it tremendously challenging to translate benevolent aims into practice' is underdetermined by the evidence presented. Sections 3.1–3.5 map social harms to technical assumptions at the level of analogy and correlation, but no counterfactual or comparative case is given. For example, the annotator exploitation and 'ethics dumping' described in Section 2.1 are plausibly driven by labor markets and outsourcing incentives, not by the likelihood factorization in Eq. (1), and the surveillance and persecution examples in Section 2.5 are plausibly driven by state and corporate power rather than by averaged or surrogate losses. The paper should either soften the causal wording to 'hypothesis' or 'barrier that deserves investigation,' or provide concrete evidence that removing or altering these technical assumptions would change the social outcomes. The limitation section (Section 4) acknowledges the analogy's frame, but it does not address this specific underdetermination.
  2. [Section 2.3] The 'superficial alignment hypothesis' is presented as a settled finding using Refs. [165] and [97], with the statement that post-training tuning only affects style and 'at best result[s] in superficial improvement.' This hypothesis is actively debated in the alignment literature, and the paper does not acknowledge competing evidence or the conditions under which alignment modifies behavior more deeply. Since this premise undergirds the 'baking process is irreversible' argument and the recommendations in Section 3.3, the presentation should be qualified to reflect the current state of evidence rather than asserted as established fact.
  3. [Section 3.2] The claim that Capsule networks underperformed 'not necessarily due to inferior design, but rather a direct function of programming frameworks being excessively tailored to homogenized deep learning recipes' overstates what the cited work [11] establishes. The cited HotOS paper argues that ML systems are hard to change for a variety of systems and software reasons, not that framework tailoring is the sole or direct cause of Capsule networks' performance gap. This strong attribution weakens an otherwise reasonable point about homogenization and should be softened or re-supported.
minor comments (3)
  1. [Section 3.4, Eq. (5)] The phrase 'Following earlier equation 2' should read 'Following Eq. (2)' for clarity, since the reference is to a numbered equation, not a section.
  2. [Section 2.1] The claim that ChatGPT's frequent use of 'delve' 'was later attributed to' Nigerian English relies on a journalistic source; it would be more precise to present this as a reported hypothesis rather than an established fact.
  3. [References] A few references are informal or non-archival, such as [2] (a glossary entry), [89] (a news article without a venue), and [132] (a magazine article). For a paper that leans heavily on citations, adding archival sources or clearly labeling media reports would strengthen reproducibility of the evidence base.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the paper's argument is analogy-based, its technical claims are supported by external evidence, and its self-citations are not load-bearing.

full rationale

This is a conceptual/position paper rather than a derivation, so the classical circularity patterns do not directly apply. The central claim (Section 3, introductory paragraph) that technical foundations make it 'tremendously challenging to translate benevolent aims into practice' is supported by mapping social outcomes to ML assumptions (i.i.d. likelihoods, homogenized architectures, single-cycle training, averaged losses, surrogate objectives), but each mapping is argued from external empirical work (e.g., Birhane et al. on LAION hate content, LIMA on superficial alignment, Blodgett et al. on benchmarks, Chekroud et al. on clinical prediction). The paper invokes self-citations (e.g., Mundt et al. on continual learning, CLEVA-Compass, Queer in AI) only as recommendations or supporting background, not as the proof of the central claim. Section 4 explicitly disclaims that social challenges will be overcome by AI design alone and acknowledges the analogy's limits, which further reduces any risk of the argument being circular. No equation in the paper is defined in terms of its conclusion, and no fitted quantity is relabeled as a prediction. Thus no specific circular reduction can be exhibited.

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

The central claim rests on the metaphor's accuracy and on several contested empirical claims that are treated as settled. There are no free parameters or invented entities.

assumptions (4)
  • domain assumption The AI cake metaphor, extended to the full lifecycle, is an accurate enough lens to map social ramifications onto technical foundations.
    The paper's entire structure depends on the parallels it draws between cake stages and AI stages (Sections 2.1-2.5). This framing is asserted, not derived.
  • domain assumption The i.i.d. assumption is a critical technical root of data-related social harms.
    Section 3.1 treats the likelihood factorization in Eq. (1) as the reason data interdependencies are neglected, linking it to opacity and exploitation. This causal chain is assumed rather than empirically demonstrated.
  • domain assumption Post-training modifications are largely superficial, per the superficial alignment hypothesis.
    Section 2.3 relies on LIMA [165] and related work to claim that fine-tuning cannot change core capabilities. This is a contested empirical claim presented as settled.
  • domain assumption Aggregate evaluation metrics are deeply rooted in the algorithmic stack and cause harm.
    Section 3.4 derives average loss from Eq. (5) and attributes fairness and representation failures to this technical choice, an interpretive assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Cake that is Intelligence and Who Gets to Bake it: An AI Analogy and its Implications for Participation." pith.science (2026). https://pith.science/paper/YJN4X3JF

@misc{pith2026250203038,
  author       = {Pith},
  title        = {Pith review of: The Cake that is Intelligence and Who Gets to Bake it: An AI Analogy and its Implications for Participation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/YJN4X3JF}},
  note         = {Machine review of arXiv:2502.03038}
}
read the original abstract

In a widely popular analogy by Turing Award Laureate Yann LeCun, machine intelligence has been compared to cake - where unsupervised learning forms the base, supervised learning adds the icing, and reinforcement learning is the cherry on top. We expand this 'cake that is intelligence' analogy from a simple structural metaphor to the full life-cycle of AI systems, extending it to sourcing of ingredients (data), conception of recipes (instructions), the baking process (training), and the tasting and selling of the cake (evaluation and distribution). Leveraging our re-conceptualization, we describe each step's entailed social ramifications and how they are bounded by statistical assumptions within machine learning. Whereas these technical foundations and social impacts are deeply intertwined, they are often studied in isolation, creating barriers that restrict meaningful participation. Our re-conceptualization paves the way to bridge this gap by mapping where technical foundations interact with social outcomes, highlighting opportunities for cross-disciplinary dialogue. Finally, we conclude with actionable recommendations at each stage of the metaphorical AI cake's life-cycle, empowering prospective AI practitioners, users, and researchers, with increased awareness and ability to engage in broader AI discourse.

Figures

Figures reproduced from arXiv: 2502.03038 by the authors.

Figure 1
Figure 1. An illustration of the AI–cake analogy. (Left) Traditionally, the cake was used to provide a structural metaphor for machine intel [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. An overview of section three’s considerations with respect to the AI-cake’s technical foundation and our future recommendations [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

167 extracted references · 51 canonical work pages

  1. [165]

    Chunting Zhou, Pengfei Liu, Puxin Xu, Srini Iyer, Jiao Sun, Yuning Mao, Xuezhe Ma, Avia Efrat, Ping Yu, Lili Yu, Susan Zhang, Gargi Ghosh, Mike Lewis, Luke Zettlemoyer, and Omer Levy. 2023. LIMA: Less is More for Alignment. Neural Information Processing Systems (NeurIPS) (2023)

    work page 2023

  2. [97]

    Bill Yuchen Lin, Abhilasha Ravichandar, Ximing Lu, Nouha Dziri, Melanie Sclar, Khyathi Chandu, Chandra Bhagavatula, and Yejin Choi. 2024. The Unlocking Spell on Base LLMs: Rethinking Alignment via In-Context Learning. International Conference on Learning Representations (ICLR) (2024)

    2024

  3. [11]

    Paul Barham and Michael Isard. 2019. Machine Learning Systems are Stuck in a Rut. Proceedings of the Workshop on Hot Topics in Operating Systems (HotOS) (2019)

    2019

  4. [1]

    WinoGrande: An Adversarial Winograd Schema Challenge at Scale

    2019. WinoGrande: An Adversarial Winograd Schema Challenge at Scale

    2019

  5. [2]

    Shirking

    2024. Shirking. https://www.nasdaq.com/glossary/s/shirking. Last accessed: 2025-01-12

    2024

  6. [3]

    Murray, Benoit Steiner, Paul Tucker, Vijay Vasudevan, Pete Warden, Martin Wicke, Yuan Yu, and Xiaoqiang Zheng

    Martín Abadi, Paul Barham, Jianmin Chen, Zhifeng Chen, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Geoffrey Irving, Michael Isard, Manjunath Kudlur, Josh Levenberg, Rajat Monga, Sherry Moore, Derek G. Murray, Benoit Steiner, Paul Tucker, Vijay Vasudevan, Pete Warden, Martin Wicke, Yuan Yu, and Xiaoqiang Zheng. 2016. TensorFlow: A system for...

    2016

  7. [4]

    Hewett, Jamie Huynh, Mojan Javaheripi, Xin Jin, Piero Kauffmann, Nikos Karampatziakis, Dongwoo Kim, Mahoud Khademi, Lev Kurilenko, James R

    Marah Abdin, Sam Ade Jacobs, Ammar Ahmad Awan, Jyoti Aneja, Ahmed Awadallah, Hany Awadalla, Nguyen Bach, Amit Bahree, Arash Bakhtiari, Jianmin Bao, Harkirat Behl, Alon Benhaim, Misha Bilenko, Johan Bjorck, Sébastien Bubeck, Qin Cai, Martin Cai, Caio César Teodoro Mendes, Weizhu Chen, Vishrav Chaudhary, Dong Chen, Dongdong Chen, Yen-Chun Chen, Yi-Ling Chen...

    arXiv 2024

  8. [5]

    Ada-Lovelace-Institute. 2021. Participatory data stewardship. Technical report available at: https://www.adalovelaceinstitute.org/report/ participatory-data-stewardship/

    2021

Show all 167 references
  1. [6]

    Abejide Ade-Ibijola and Chinedu Okonkwo. 2023. Artificial Intelligence in Africa: Emerging Challenges . Springer International Publishing, 101–117

    2023

  2. [7]

    Marcin Andrychowicz, Filip Wolski, Alex Ray, Jonas Schneider, Rachel Fong, Peter Welinder, Bob McGrew, Josh Tobin, Pieter Abeel, and Wojciech Zaremba. 2017. Hindsight Experience Replay. Neural Information Processing Systems (NeurIPS) (2017)

    2017

  3. [8]

    Emmanuel Ogiemwonyi Arakpogun, Ziad Elsahn, Femi Olan, and Farid Elsahn. 2021. Artificial Intelligence in Africa: Challenges and Opportunities . Springer International Publishing, 375–388

    2021

  4. [9]

    Arnstein

    Sherry R. Arnstein. 1969. A ladder of citizen participation. Journal of the American Institute of planners 35 (1969), 216–224. Issue 4

    1969

  5. [10]

    Timur Ash. 1989. Dynamic Node Creation in Backpropagation Networks. Connection Science 1, 4 (1989), 365–375

    1989

  6. [12]

    R. K. E. Bellamy, K. Dey, M. Hind, S. C. Hoffman, S. Houde, K. Kannan, P. Lohia, J. Martino, S. Mehta, A. Mojsilović, S. Nagar, K. Natesan Ramamurthy, J. Richards, D. Saha, P. Sattigeri, M. Singh, K. R. Varshney, and Y. Zhang. 2019. AI Fairness 360: An extensible toolkit for d...

    2019

  7. [13]

    Bender and Batya Friedman

    Emily M. Bender and Batya Friedman. 2018. Data Statements for Natural Language Processing: Toward Mitigating System Bias and Enabling Better Science. Transactions of the Association for Computational Linguistics 6 (2018), 587–604

    2018

  8. [14]

    Stevie Bergman, Lisa Anne Hendricks, Maribeth Rauh, Boxi Wu, William Agnew, Markus Kunesch, Isabella Duan, Iason Gabriel, , and William Isaac

    A. Stevie Bergman, Lisa Anne Hendricks, Maribeth Rauh, Boxi Wu, William Agnew, Markus Kunesch, Isabella Duan, Iason Gabriel, , and William Isaac. 2023. Representation in AI Evaluations. Proceedings of the 2023 ACM Conference on Fairness, Accountability, and Transparency (FAccT...

    2023

  9. [15]

    James Betker, Gabriel Goh, Li Jing, Tim Brooks, Jianfeng Wang, Linjie Li, Long Ouyang, Juntang Zhuang, Joyce Lee, Yufei Guo, Wesam Manassra, Prafulla Dhariwal, Casey Chu, Yunxin Jiao, and Aditya Ramesh. [n. d.]. OpenAI report: ImprovingImageGenerationwithBetterCaptions

  10. [16]

    R. Binns. 2020. On the apparent conflict between individual and group fairness. Proceedings of the Conference on Fairness, Accountability and Transparency (FAccT) (2020)

    2020

  11. [17]

    Abeba Birhane. 2022. The unseen Black faces of AI algorithms. Nature 610 (2022), 451–452

    2022

  12. [18]

    Abeba Birhane, William Isaac, Vinodkumar Prabhakaran, Mark Diaz, Madeleine Clare Elish, Iason Gabriel, and Shakir Mohamed. 2022. Power to the People? Opportunities and Challenges for Participatory AI. ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimizat...

    2022

  13. [19]

    Abeba Birhane, Vinay Prabhu, Sang Han, Vishnu Naresh Boddeti, and Alexandra Sasha Luccioni. 2023. Into the LAIONs Den: Investigating Hate in Multimodal Datasets. Neural Information Processing Systems (NeurIPS) (2023)

    2023

  14. [20]

    Christoper M. Bishop. 2006. Pattern Recognition and Machine Learning . Springer

    2006

  15. [21]

    Su Lin Blodgett, Gilsinia Lopez, Alexandra Olteanu, Robert Sim, and Hanna Wallach. 2021. Stereotyping Norwegian Salmon: An Inventory of Pitfalls in Fairness Benchmark Datasets. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th...

    2021

  16. [22]

    Rishi Bommasani. 2022. Evaluation for Change. Preprint arXiv:2212.11670 (2022). 16 Mundt et al

    2022 arXiv

  17. [23]

    Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S

    Rishi Bommasani, Drew A. Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S. Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill, Erik Brynjolfsson, S. Buch, Dallas Card, Rodrigo Castellon, Niladri S. Chatterji, Annie S. Chen, Kathleen A. Creel, ...

    2021 arXiv

  18. [24]

    Ali Borji. 2023. A Categorical Archive of ChatGPT Failures. Preprint arXiv:2302.03494 (2023)

    2023 arXiv

  19. [25]

    Boult, Steve Cruz, Akshay R

    Terrance E. Boult, Steve Cruz, Akshay R. Dhamija, Manuel Gunther, James Henrydoss, and Walter J. Scheirer. 2019. Learning and the Unknown : Surveying Steps Toward Open World Recognition. AAAI Conference on Artificial Intelligence (AAAI) (2019)

    2019

  20. [26]

    James Bradbury, Roy Frostig, Peter Hawkins, Matthew James Johnson, Chris Leary, Dougal Maclaurin, George Necula, Adam Paszke, Jake VanderPlas, Skye Wanderman-Milne, and Qiao Zhang. 2018. JAX: composable transformations of Python+NumPy programs: http://github.com/google/jax

    2018

  21. [27]

    Tim Brooks, Bill Peebles, Connor Holmes, Will DePue, Yufei Guo, Li Jing, David Schnurr, Joe Taylor, Troy Luhman, Eric Luhman, Clarence Ng, Ricky Wang, and Aditya Ramesh. 2024. Video Generation Models as World Simulators. https://openai.com/index/video-generation-models-as- wor...

    2024

  22. [28]

    Meredith Broussard. 2019. Artificial Unintelligence. The MIT Press

    2019

  23. [29]

    Meredith Broussard. 2023. More than a Glitch: Confronting Race, Gender, and Ability Bias in Tech . MIT Press

    2023

  24. [30]

    Joy Buolamwini and Timnit Gebru. 2018. Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification. Proceedings of the Conference on Fairness, Accountability and Transparency (FAccT) (2018)

    2018

  25. [31]

    Burnell, W

    R. Burnell, W. Schellaert, J. Burden, T. D. Ullman, F. Martinez-Plumed, J. B. Tenenbaum, D. Rutar, L. G. Cheke, J. Sohl-Dickstein, M. Mitchell, D. Kiela, M. Shanahan, E. M. Voorhees, A. G. Cohn, J. Z. Leibo, and J. Hernandez-Orallo. 2023. Rethink reporting of evaluation result...

    2023

  26. [32]

    Florian Peter Busch, Roshni Kamath, Rupert Mitchell, Wolfgang Stammer, Kristian Kersting, and Martin Mundt. 2024. Where is the Truth? The Risk of Getting Confounded in a Continual World. arXiv preprint arXiv:2402.06434 (2024)

    2024 arXiv

  27. [33]

    Cadieu, Ha Hong, Daniel L.K

    Charles F. Cadieu, Ha Hong, Daniel L.K. Yamins, Nicolas Pinto, Diego Ardila, Ethan A. Solomon, Najib J. Majaj, and James J. DiCarlo. 2014. Deep Neural Networks Rival the Representation of Primate IT Cortex for Core Visual Object Recognition. PLoS Computational Biology 10, 12 (2014)

    2014

  28. [34]

    Andrew Carlson, Justin Betteridge, and Bryan Kisiel. 2010. Toward an Architecture for Never-Ending Language Learning. AAAI Conference on Artificial Intelligence (AAAI) (2010)

    2010

  29. [35]

    Alessandro Castelnovo, Riccardo Crupi, Greta Greco, Daniele Regoli, Ilaria Giuseppina Penco, and Andrea Claudio Cosentini. 2022. A clarification of the nuances in the fairness metrics landscape. Nature Scientific Reports 12 (2022)

    2022

  30. [36]

    A. M. Chekroud, M. Hawrilenko, H. Loho, J. Bondar, R. Gueorguieva, A. Hasan, J. Kambeitz, P. R. Corlett, N. Koutsouleris, H. M. Krumholz, J. H. Krystal, and M. Paulus. 2024. Illusory generalizability of clinical prediction models. Science 383 (2024)

    2024

  31. [37]

    Lingjiao Chen, Matei Zaharia, and James Zou. 2023. How is ChatGPT’s behavior changing over time? arXiv preprint arXiv:2307.09009 (2023)

    2023 arXiv

  32. [38]

    Xinlei Chen, Abhinav Shrivastava, and Abhinav Gupta. 2013. NEIL: Extracting visual knowledge from web data.Proceedings of the IEEE International Conference on Computer Vision (CVPR) (2013)

    2013

  33. [39]

    Zhiyuan Chen and Bing Liu. 2018. Lifelong machine learning. Synthesis Lectures on Artificial Intelligence and Machine Learning 12, 3 (2018), 1–207

    2018

  34. [40]

    Jennifer Chien and David Danks. 2024. Beyond Behaviorist Representational Harms: A Plan for Measurement and Mitigation. Proceedings of the Conference on Fairness, Accountability and Transparency (FAccT) (2024), 933–946

    2024

  35. [41]

    Eunsol Choi, He He, Mohit Iyyer, Mark Yatskar, Wen-tau Yih, Yejin Choi, Percy Liang, and Luke Zettlemoyer. 2018. QuAC: Question Answering in Context. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing . Association for Computational Lingu...

    2018 doi

  36. [42]

    Christopher Clark, Kenton Lee, Ming-Wei Chang, Tom Kwiatkowski, Michael Collins, and Kristina Toutanova. 2019. BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions. In NAACL

    2019

  37. [43]

    Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, Christopher Hesse, and John Schulman. 2021. Training Verifiers to Solve Math Word Problems. arXiv preprint arXiv:2110.14168 (2021)

    2021 arXiv

  38. [44]

    Ronan Collobert, Koray Kavukcuoglu, and Clément Farabet. 2011. Torch7: A matlab-like environment for machine learning. Neural Information Processing Systems (NeurIPS), BigLearn Workshop (2011). The Cake that is Intelligence and Who Gets to Bake it: An AI Analogy and its Implic...

    2011

  39. [45]

    Gaebler, Hamed Nilforoshan, Ravi Shroff, and Sharad Goel

    Sam Corbett-Davies, Johann D. Gaebler, Hamed Nilforoshan, Ravi Shroff, and Sharad Goel. 2023. The Measure and Mismeasure of Fairness.Journal of Machine Learning Research (JMLR) 24 (2023), 1–117

    2023

  40. [46]

    Trevor Darrell, Marius Kloft, Massimiliano Pontil, Gunnar Rätsch, and Erik Rodner. 2015. Machine Learning with Interdependent and Non-identically Distributed Data (Dagstuhl Seminar 15152). Dagstuhl Reports (2015), 18–55

    2015

  41. [47]

    Add Diverse Stakeholders and Stir

    Fernando Delgado, Stephen Yang, Michael Madaio, and Qian Yang. 2021. Stakeholder Participation in AI: Beyond "Add Diverse Stakeholders and Stir". Human-Centered AI workshop at NeurIPS (2021)

    2021

  42. [48]

    Fernando Delgado, Stephen Yang, Michael Madaio, and Qian Yang. 2023. The Participatory Turn in AI Design: Theoretical Foundations and the Current State of Practice. ACM Conference on Equity and Access in Algorithms, Mechanisms, and Optimization (EAAMO) (2023)

    2023

  43. [49]

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. 2009. ImageNet: A large-scale hierarchical image database. Computer Vision and Pattern Recognition (CVPR) (2009)

    2009

  44. [50]

    Emily Denton, Alex Hanna, Razvan Amironesei, Andrew Smart, and Hilary Nicole. 2021. On the Genealogy of Machine Learning Datasets: A Critical History of ImageNet. Big Data and Society (2021)

    2021

  45. [51]

    Sunipa Dev, Masoud Monajatipoor, Anaelia Ovalle, Arjun Subramonian, Jeff M Phillips, and Kai-Wei Chang. 2021. Harms of gender exclusivity and challenges in non-binary representation in language technologies. arXiv preprint arXiv:2108.12084 (2021)

    2021 arXiv

  46. [52]

    Murat Dundar, Balaji Krishnapuram, Jinbo Bi, and R Bharat Rao. 2007. Learning Classifiers When the Training Data Is Not IID. In Proceedings of the 20th International Joint Conference on Artificial Intelligence (IJCAI)

    2007

  47. [53]

    European-Commission. 2015. Horizon 2020 Work Programme 2014-2015, Science with and for Society. https://research-and-innovation.ec.europa. eu/funding/funding-opportunities/funding-programmes-and-open-calls/horizon-2020_en

    2015

  48. [54]

    Utku Evci, Bart van Merrienboer, Thomas Unterthiner, Fabian Pedregosa, and Max Vladymyrov. 2022. GradMax: Growing Neural Networks using Gradient Information. International Conference on Learning Representations (ICLR) (2022)

    2022

  49. [55]

    Fabian Ferrari, Jose van Dijck, and Antal van den Bosch. 2023. Foundation models and the privatization of public knowledge. Nature machine intelligence 5 (2023)

    2023

  50. [56]

    Timo Flesch, Jan Balaguer, Ronald Dekker, Hamed Nili, and Christopher Summerfield. 2018. Comparing continual task learning in minds and machines. Proceedings of the National Academy of Sciences of the United States of America PNAS 115 (2018). Issue 44

    2018

  51. [57]

    Nagy, Andrew Saxe, and Christopher Summerfield

    Timo Flesch, David G. Nagy, Andrew Saxe, and Christopher Summerfield. 2023. Modelling continual learning in humans with Hebbian context gating and exponentially decaying task signals. PLoS Computational Biology 19, 1 (2023)

    2023

  52. [58]

    Luciano Floridi. 2019. Translating Principles into Practices of Digital Ethics: Five Risks of Being Unethical. Philosophy and Technology 32 (2019), 185–193

    2019

  53. [59]

    Robert M. French. 1992. Semi-distributed Representations and Catastrophic Forgetting in Connectionist Networks. Connection Science 4, 3-4 (1992), 365–377

    1992

  54. [60]

    Felix Friedrich, Katharina Hämmerl, Patrick Schramowski, Manuel Brack, Jindrich Libovicky, Kristian Kersting, and Alexander Fraser. 2024. Multilingual Text-to-Image Generation Magnifies Gender Stereotypes and Prompt Engineering May Not Help You. Preprint arXiv:2401.16092 (2024)

    2024 arXiv

  55. [61]

    Timnit Gebru, Jamie Morgenstern, Briana Vecchione, Jennifer Wortman Vaughan, Hanna Wallach, Hal Daumé, and Kate Crawford. 2018. Datasheets for Datasets. Preprint arXiv:1803.09010 (2018)

    2018 arXiv

  56. [62]

    Avijit Ghosh, Cedric Whitney, Yacine Jernite, and Irene Solaiman. 2024. Social Impact Dashboard. https://huggingface.co/spaces/evijit/ SIMPDashboard. Accessed: 2025-01-07

    2024

  57. [63]

    Chris Gilliard. 2024. The Deeper Problem With Google’s Racially Diverse Nazis: Generative AI is not built to honestly mirror reality, no matter what its creators say. The Atlantic: https://www.theatlantic.com/technology/archive/2024/02/google-gemini-diverse-nazis/677575/

    2024

  58. [64]

    Goodfellow, Yoshua Bengio, and Aaron Courville

    Ian J. Goodfellow, Yoshua Bengio, and Aaron Courville. 2016. Deep Learning. MIT Press

    2016

  59. [65]

    Google. 2024. Google Gemini. https://gemini.google.com

    2024

  60. [66]

    Google-Cloud. 2023. Google Cloud guides: ML solutions overview, the ML workflow. https://cloud.google.com/ai-platform/docs/ml-solutions- overview. 2024-03-20

    2023

  61. [67]

    Charles G. Gross. 2000. Neurogenesis in the adult brain: Death of a dogma. Nature Reviews Neuroscience 1, 1 (2000), 67–73

    2000

  62. [68]

    Raia Hadsell, Dushyant Rao, Andrei A Rusu, and Razvan Pascanu. 2020. Embracing change: Continual learning in deep neural networks. Trends in cognitive sciences 24, 12 (2020), 1028–1040

    2020

  63. [69]

    Shuang Hao, Wenfeng Han, Tao Jiang, Yiping Li, Haonan Wu, Chunlin Zhong, Zhangjun Zhou, and He Tang. 2024. Synthetic Data in AI: Challenges, Applications, and Ethical Implications. Preprint arXiv:2401.01629 (2024)

    2024 arXiv

  64. [70]

    Drew Harwell and Eva Dou. 2020. Huawei tested AI software that could recognize Uighur minorities and alert police

    2020

  65. [71]

    frontier AI

    Gina Helfrich. 2024. The harms of terminology: why we should reject so-called “frontier AI”. AI and Ethics (2024)

    2024

  66. [72]

    Dan Hendrycks and Kevin Gimpel. 2017. A baseline for detecting misclassified and out-of-distribution examples in neural networks. International Conference on Learning Representations (ICLR) (2017)

    2017

  67. [73]

    Ferdinand A. Hermens. 1958. The Tyranny of the Majority. Social Research 25 (1958), 37–52

    1958

  68. [74]

    Alex Hern. 2024. TechScape: How cheap, outsourced labour in Africa is shaping AI English. The Guardian: https://www.theguardian.com/ technology/2024/apr/16/techscape-ai-gadgest-humane-ai-pin-chatgpt. 18 Mundt et al

    2024

  69. [75]

    Dominik Hintersdorf, Lukas Struppek, Manuel Brack, Felix Friedrich, Patrick Schramowski, and Kristian Kersting. 2022. Does CLIP Know my Face? Preprint arXiv:2209.07341 (2022)

    2022 arXiv

  70. [76]

    Kurt Hornik, Maxwell Stinchcombe, and Halbert White. 1989. Multilayer feedforward networks are universal approximators. Neural Networks (1989)

    1989

  71. [77]

    The White House. 2023. FACT SHEET: President Biden Issues Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence. https: //resourcecenter.ieee.org/government/usa/cisgovph0110. Last accessed: 2025-02-02

    2023

  72. [78]

    Marcus Hutter. 2022. Testing Independence of Exchangeable Random Variables. preprint arXiv:2210.12392 (2022)

    2022 arXiv

  73. [79]

    Mojan Javaheripi and Sebastien Bubeck. 2023. Phi-2 the surprising power of small language models. Microsoft Blog: https://www.microsoft.com/en- us/research/blog/phi-2-the-surprising-power-of-small-language-models/

    2023

  74. [80]

    Gomez, Noam Shazeer, Ashish Vaswani, Niki Parmar, Llion Jones, and Jakob Uszkoreit

    Lukasz Kaiser, Aidan N. Gomez, Noam Shazeer, Ashish Vaswani, Niki Parmar, Llion Jones, and Jakob Uszkoreit. 2017. One Model To Learn Them All. Prerint arXiv:1706.05137 (2017)

    2017 arXiv

  75. [81]

    Pratyusha Ria Kalluri, William Agnew, Myra Cheng, Kentrell Owens, Luca Soldaini, and Abeba Birhane. 2023. The Surveillance AI Pipeline. Preprint arXiv:2309.15084 (2023)

    2023 arXiv

  76. [82]

    Remi Kazmierczak, Gianni Franchi, Nacim Belkhir, Antoine Manzanera, and David Filliat. 2022. A Study of Deep Perceptual Metrics for Image Quality Assessment. arXiv preprint arXiv:2202.08692 (2022)

    2022 arXiv

  77. [83]

    Kingma and Jimmy Ba

    Diederik P. Kingma and Jimmy Ba. 2015. Adam: A Method for Stochastic Optimization.International Conference on Learning Representations (2015)

    2015

  78. [84]

    Diederik P Kingma and Max Welling. 2014. Auto-Encoding Variational Bayes. In 2nd International Conference on Learning Representations, ICLR 2014, Banff, AB, Canada, April 14-16, 2014, Conference Track Proceedings

    2014

  79. [85]

    Diederik P Kingma and Max Welling. 2019. An Introduction to Variational Autoencoders. Foundations and Trends in Machine Learning (2019)

    2019

  80. [86]

    Hannah Rose Kirk, Alexander Whitefield, Paul Röttger, Andrew Bean, Katerina Margatina, Juan Ciro, Rafael Mosquera, Max Bartolo, Adina Williams, He He, et al. 2024. The PRISM Alignment Project: What Participatory, Representative and Individualised Human Feedback Reveals About t...

    2024 arXiv

  81. [87]

    Jon Kleinberg, Sendhil Mullainathan, and Manish Raghavan. 2017. Inherent Trade-Offs in the Fair Determination of Risk Scores. Proceedings of Innovations in Theoretical Computer Science (ITCS) (2017)

    2017

  82. [88]

    John Koetsier. 2023. ChatGPT Burns Millions Every Day. Can Computer Scientists Make AI One Million Times More Efficient? Forbes: https://www.forbes.com/sites/johnkoetsier/2023/02/10/chatgpt-burns-millions-every-day-can-computer-scientists-make-ai-one-million- times-more-effici...

    2023

  83. [89]

    Leo Komminoth. 2023. Can AI address Africa’s agricultural trade deficit?

    2023

  84. [90]

    Agustinus Kristiadi. 2016. KL Divergence: Forward vs Reverse? https://agustinus.kristia.de/blog/forward-reverse-kl/

    2016

  85. [91]

    Dhireesha Kudithipudi, Mario Aguilar-Simon, Jonathan Babb, Maxim Bazhenov, Douglas Blackiston, Josh Bongard, Andrew P Brna, Suraj Chakravarthi Raja, Nick Cheney, Jeff Clune, and Others. 2022. Biological underpinnings for lifelong learning machines. Nature Machine Intelligence ...

    2022

  86. [92]

    M. J. Kusner, J. Loftus, C. Russell, and R. Silva. 2017. Counterfactual Fairness. Neural Information Processing Systems (NeurIPS) (2017)

    2017

  87. [93]

    Sebastian Lapuschkin, Stephan Wäldchen, Alexander Binder, Grégoire Montavon, Wojciech Samek, and Klaus-Robert Müller. 2019. Unmasking Clever Hans predictors and assessing what machines really learn. Nature Communications 10, 1 (2019)

    2019

  88. [94]

    Yann A LeCun, Léon Bottou, Genevieve B Orr, and Klaus-Robert Müller. 2012. Efficient BackProp. In Neural Networks: Tricks of the Trade: Second Edition. Springer Berlin Heidelberg, Berlin, Heidelberg, 9–48

    2012

  89. [95]

    Tian Li, Ahmad Beirami, Maziar Sanjabi, and Virginia Smith. 2023. On tilted losses in machine learning: theory and applications. Journal of Machine Learning Research (2023)

    2023

  90. [96]

    Hunter Lightman, Vineet Kosaraju, Yura Burda, Harri Edwards, Bowen Baker, Teddy Lee, Jan Leike, John Schulman, Ilya Sutskever, and Karl Cobbe. 2023. Let’s Verify Step by Step. arXiv preprint arXiv:2305.20050 (2023)

    2023 arXiv

  91. [98]

    Chin-Yew Lin. 2004. ROUGE: A package for automatic evaluation of summaries. In Text Summarization Branches Out

    2004

  92. [99]

    Lipton, Alexandra Chouldechova, and Julian McAuley

    Zachary C. Lipton, Alexandra Chouldechova, and Julian McAuley. 2019. Does mitigating ML’s impact disparity require treatment disparity? Neural Information Processing Systems (NeurIPS) (2019)

    2019

  93. [100]

    Kasper Groes Albin Ludvigsen. 2023. The carbon footprint of GPT-4. https://towardsdatascience.com/the-carbon-footprint-of-gpt-4-d6c676eb21ae. Last accessed: 2025-02-02

    2023

  94. [101]

    Kristian Lum, Yunfeng Zhang, and Amanda Bower. 2022. De-biasing “bias” measurement. Proceedings of the 2022 ACM Conference on Fairness, Accountability, and Transparency (FAccT) (2022), 379–389

    2022

  95. [102]

    Michael McCloskey and Neal J. Cohen. 1989. Catastrophic Interference in Connectionist Networks : The Sequential Learning Problem. Psychology of Learning and Motivation - Advances in Research and Theory 24, C (1989), 109–165

    1989

  96. [103]

    Rupert Mitchell, Robin Menzenbach, Kristian Kersting, and Martin Mundt. 2024. Self Expanding Neural Networks. Preprint arXiv:2307.04526 (2024)

    2024 arXiv

  97. [104]

    The Cake that is Intelligence and Who Gets to Bake it: An AI Analogy and its Implications for Participation 19

    T Mitchell, W Cohen, E Hruschka, P Talukdar, B Yang, J Betteridge, A Carlson, B Dalvi, M Gardner, B Kisiel, J Krishnamurthy, N Lao, K Mazaitis, T Mohamed, N Nakashole, E Platanios, A Ritter, M Samadi, B Settles, R Wang, D Wijaya, A Gupta, X Chen, A Saparov, M Greaves, and J We...

  98. [105]

    Martin Mundt, Yongwon Hong, Iuliia Pliushch, and Visvanathan Ramesh. 2023. A wholistic view of continual learning with deep neural networks: Forgotten lessons and the bridge to active and open world learning. Neural Networks 160 (2023), 306–336

    2023

  99. [106]

    Martin Mundt, Steven Lang, Quentin Delfosse, and Kristian Kersting. 2022. CLEVA-Compass: A Continual Learning Evaluation Assessment Compass to Promote Research Transparency and Comparability. International Conference on Learning Representations (ICLR) (2022)

    2022

  100. [107]

    Kevin Patrick Murphy. 2012. Machine Learning: A Probabilistic Perspective . MIT Press

    2012

  101. [108]

    Esther Mwema and Abeba Birhane. 2024. Undersea cables in Africa: The new frontiers of digital colonialism. First Monday 29 (2024). Issue 4

    2024

  102. [109]

    Preetum Nakkiran, Gal Kaplun, Yamini Bansal, Tristan Yang, Boaz Barak, and Ilya Sutskever. 2020. Deep Double Descent: Where Bigger Models and More Data Hurt. International Conference on Learning Representations (ICLR) (2020)

    2020

  103. [110]

    Teh, Dilan Gorur, and Balaji Lakshminarayanan

    Eric Nalisnick, Akihiro Matsukawa, Yee W. Teh, Dilan Gorur, and Balaji Lakshminarayanan. 2019. Do Deep Generative Models Know What They Don’t Know? International Conference on Learning Representations (ICLR) (2019)

    2019

  104. [111]

    Narayanan

    A. Narayanan. 2018. Translation tutorial: 21 fairness definitions and their politics. Proceedings of the Conference on Fairness, Accountability and Transparency (FAccT) (2018)

    2018

  105. [112]

    Safiya Umoja Noble. 2018. Algorithms of Oppression: How Search Engines Reinforce Racism . NYU Press

    2018

  106. [113]

    Noyb. 2024. ChatGPT provides false information about people, and OpenAI can’t correct it. Noyb: https://noyb.eu/en/chatgpt-provides-false- information-about-people-and-openai-cant-correct-it

    2024

  107. [114]

    Deparment of Commerce

    U.S. Deparment of Commerce. 2025. Biden-Harris Administration Announces Regulatory Framework for the Responsible Diffusion of Advanced Artificial Intelligence Technology. https://www.bis.gov/press-release/biden-harris-administration-announces-regulatory-framework-responsible- ...

    2025

  108. [115]

    German Federal Ministry of Labour and Social Affairs. 2023. Supply Chain Act. https://www.bmas.de/EN/Europe-and-the-World/International/ Supply-Chain-Act/supply-chain-act.html

    2023

  109. [116]

    Sculley, Sebastian Nowozin, Joshua V

    Yaniv Ovadia, Emily Fertig, Jie Ren, Zachary Nado, D. Sculley, Sebastian Nowozin, Joshua V. Dillon, Balaji Lakshminarayanan, and Jasper Snoek

  110. [117]

    I’m fully who I am

    Anaelia Ovalle, Palash Goyal, Jwala Dhamala, Zachary Jaggers, Kai-Wei Chang, Aram Galstyan, Richard Zemel, and Rahul Gupta. 2023. “I’m fully who I am”: Towards Centering Transgender and Non-Binary Voices to Measure Biases in Open Language Generation. InProceedings of the 2023 ...

    2023

  111. [118]

    Anaelia Ovalle, Krunoslav Lehman Pavasovic, Louis Martin, Luke Zettlemoyer, Eric Michael Smith, Adina Williams, and Levent Sagun. 2024. The Root Shapes the Fruit: On the Persistence of Gender-Exclusive Harms in Aligned Language Models. arXiv preprint arXiv:2411.03700 (2024)

    2024 arXiv

  112. [119]

    Pan and Qiang Yang

    Sinno J. Pan and Qiang Yang. 2010. A Survey on Transfer Learning. IEEE Transactions on Knowledge and Data Engineering (TKDE) 22, 10 (2010)

    2010

  113. [120]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. BLEU: a method for automatic evaluation of machine translation. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics

    2002

  114. [121]

    Adam Paszke, Sam Gross, Francisco Massa, Adam Lerer, James Bradbury, Gregory Chanan, Trevor Killeen, Zeming Lin, Natalia Gimelshein, Luca Antiga, Alban Desmaison, Andreas Köpf, Edward Yang, Zach DeVito, Martin Raison, Alykhan Tejani, Sasank Chilamkurthy, Benoit Steiner, Lu Fan...

    2019

  115. [122]

    David Patterson, Joseph Gonzalez, Quoc Le, Chen Liang, Lluis-Miquel Munguia, Daniel Rothchild, David So, Maud Texier, and Jeff Dean. 2021. Carbon Emissions and Large Neural Network Training. Preprint arXiv:2104.10350 (2021)

    2021 arXiv

  116. [123]

    Judea Pearl. 2009. Causality. Cambridge Univrsity Press

    2009

  117. [124]

    Diana Benavides Prado and Patricia Riddle. 2022. A Theory for Knowledge Transfer in Continual Learning. In Conference on Lifelong Learning Agents

    2022

  118. [125]

    Organizers Of Queerinai, Anaelia Ovalle, Arjun Subramonian, Ashwin Singh, Claas Voelcker, Danica J. Sutherland, Davide Locatelli, Eva Breznik, Filip Klubicka, Hang Yuan, Hetvi J, Huan Zhang, Jaidev Shriram, Kruno Lehman, Luca Soldaini, Maarten Sap, Marc Peter Deisenroth, Maria...

    2023

  119. [126]

    Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, and Others. 2019. Language models are unsupervised multitask learners. OpenAI Blog: https:// github.com/ openai/ gpt-2 (2019)

    2019

  120. [127]

    Prabhakar Raghavan. 2024. Gemini image generation got it wrong. We’ll do better. Google Blog: https://blog.google/products/gemini/gemini- image-generation-issue/

    2024

  121. [128]

    Pranav Rajpurkar, Jian Zhang, Konstantin Lopyrev, and Percy Liang. 2016. SQuAD: 100,000+ Questions for Machine Comprehension of Text. In Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing , Jian Su, Kevin Duh, and Xavier Carreras (Eds.). Ass...

    2016 arXiv

  122. [129]

    Roger Ratcliff. 1990. Connectionist Models of Recognition Memory: Constraints Imposed by Learning and Forgetting Functions. Psychological Review 97, 2 (1990), 285–308

    1990

  123. [130]

    Scott Reed, Konrad Zolna, Emilio Parisotto, Sergio Gómez Colmenarejo, Alexander Novikov, Gabriel Barth-maron, Mai Giménez, Yury Sulsky, Jackie Kay, Jost Tobias Springenberg, Tom Eccles, Jake Bruce, Ali Razavi, Ashley Edwards, Nicolas Heess, Yutian Chen, Raia Hadsell, Oriol Vin...

    2022

  124. [131]

    Adi Robertson. 2024. Google apologizes for missing the mark after Gemini generated racially diverse Nazis. The Verge: https://www.theverge. com/2024/2/21/24079371/google-ai-gemini-generative-inaccurate-historical

    2024

  125. [132]

    Reece Rogers. 2024. Here’s How Generative AI Depicts Queer People. Wired: https://www.wired.com/story/artificial-intelligence-lgbtq- representation-openai-sora/. Last accessed: 2025-02-02

    2024

  126. [133]

    Lillicrap, and Greg Wayne

    David Rolnick, Arun Ahuja, Jonathan Schwarz, Timothy P. Lillicrap, and Greg Wayne. 2019. Experience Replay for Continual Learning. Neural Information Processing Systems (NeurIPS) (2019)

    2019

  127. [134]

    Niamh Rowe. 2023. It’s destroyed me completely: Kenyan moderators decry toll of training of AI models. The Guardian: https://www.theguardian. com/technology/2023/aug/02/ai-chatbot-training-human-toll-content-moderator-meta-openai

    2023

  128. [135]

    Rumelhart, Geoffrey E

    David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams. 1986. Learning representations by back-propagating errors. Nature 323 (1986), 533–536

    1986

  129. [136]

    Sara Sabour, Nicholas Frosst, and Geoffrey E Hinton. 2017. Dynamic routing between capsules. Advances in Neural Information Processing Systems (NeurIPS) (2017)

    2017

  130. [137]

    Saxe, Yamini Bansal, Joel Dapello, Madhu Advani, Artemy Kolchinsky, Brendan D

    Andrew M. Saxe, Yamini Bansal, Joel Dapello, Madhu Advani, Artemy Kolchinsky, Brendan D. Tracey, and David D. Cox. 2018. On the Information Bottleneck Theory of Deep Learning. International Conference on Learning Representations (ICLR) (2018)

    2018

  131. [138]

    Schölkopf*, F

    B. Schölkopf*, F. Locatello*, S. Bauer, N. R. Ke, N. Kalchbrenner, A. Goyal, and Y. Bengio. 2021. Toward Causal Representation Learning. Proc. IEEE (2021)

    2021

  132. [139]

    Doris Schroeder, Julie Cook, François Hirsch, Solveig Fenet, and Vasantha Muthuswamy. 2018. Ethics Dumping: Case Studies from North-South Research Collaborations. SpringerBriefs in Research and Innovation Governance (2018)

    2018

  133. [140]

    Christoph Schuhmann, Romain Beaumont, Richard Vencu, Cade Gordon, Ross Wightman, Mehdi Cherti, Theo Coombes, Aarush Katta, Clayton Mullis, Mitchell Wortsman, and Others. 2022. Laion-5b: An open large-scale dataset for training next generation image-text models. Advances in Neu...

    2022

  134. [141]

    Christoph Schuhmann, Richard Vencu, Romain Beaumont, Robert Kaczmarczyk, Clayton Mullis, Aarush Katta, Theo Coombes, Jenia Jitsev, and Aran Komatsuzaki. 2021. LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs. Data Centric AI NeurIPS Workshop (2021)

    2021

  135. [142]

    Ravid Schwartz-Ziv and Naftali Tishby. 2017. Opening the black box of Deep Neural Networks via Information. Preprint arXiv: 1703.00810 (2017)

    2017 arXiv

  136. [143]

    Innovation Secretary of State for Science and Technology by Command of His Majesty. 2023. Policy paper: A pro-innovation approach to AI regulation. https://www.gov.uk/government/publications/ai-regulation-a-pro-innovation-approach/white-paper. Last accessed: 2025-01-21

    2023

  137. [144]

    So, Chen Liang, and Quoc V

    David S. So, Chen Liang, and Quoc V. Le. 2019. The Evolved Transformer. International Conference on Machine Learning (ICML) (2019)

    2019

  138. [145]

    Valerie Strauss. 2018. Stephen Hawking famously said, ‘Intelligence is the ability to adapt to change. ’ But did he really say it? The Washington Post: https://www.washingtonpost.com/news/answer-sheet/wp/2018/03/29/stephen-hawking-famously-said-intelligence-is-the-ability-to-a...

    2018

  139. [146]

    Emma Strubell, Ananya Ganesh, and Andrew McCallum. 2019. Energy and Policy Considerations for Deep Learning in NLP. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics (2019), 3645–3650

    2019

  140. [147]

    Chen Sun, Abhinav Shrivastava, Saurabh Singh, and Abhinav Gupta. 2017. Revisiting Unreasonable Effectiveness of Data in Deep Learning Era. Proceedings of the IEEE International Conference on Computer Vision (ICCV) (2017)

    2017

  141. [148]

    Alon Talmor, Jonathan Herzig, Nicholas Lourie, and Jonathan Berant. 2019. CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human ...

    2019 arXiv

  142. [149]

    The Theano Development Team, Rami Al-Rfou, Guillaume Alain, Amjad Almahairi, Christof Angermueller, Dzmitry Bahdanau, Nicolas Ballas, Frédéric Bastien, Justin Bayer, Anatoly Belikov, Alexander Belopolsky, Yoshua Bengio, Arnaud Bergeron, James Bergstra, Valentin Bisson, Josh Bl...

    2016 arXiv

  143. [150]

    Sebastian Thrun. 1996. Is Learning The n-th Thing Any Easier Than Learning The First? Neural Information Processing Systems (NeurIPS) (1996)

    1996

  144. [151]

    Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, and Others. 2023. Llama 2: Open foundation and fine-tuned chat models. Preprint arXiv:2307.09288 (2023)

    2023 arXiv

  145. [152]

    Y. Z. Tsypkin. 1966. Adaptation, training and self-organization automatic control systems. A vtomatika I Telemekhanika27 (1966), 23–61

    1966

  146. [153]

    Baish, and Mike Boyle

    Meltem Sönmez Turan, Elaine Barker, John Kelsey, Kerry McKay, Mary L. Baish, and Mike Boyle. 2018. Recommendation for the Entropy Sources Used for Random Bit Generation. NIST Special Publication 800-90B (2018). https://doi.org/10.6028/NIST.SP.800-90B

    2018 doi

  147. [154]

    Vadodaria and Sebastian Jessberger

    Krishna C. Vadodaria and Sebastian Jessberger. 2014. Functional neurogenesis in the adult hippocampus: Then and now. Frontiers in Neuroscience 8 (2014)

    2014

  148. [155]

    Jonathan Vanian and Kif Leswing. 2023. ChatGPT and generative AI are blooming, but the costs can be extraordinary. CNBC: https://www.cnbc. com/2023/03/13/chatgpt-and-generative-ai-are-booming-but-at-a-very-expensive-price.html. Last accessed: 2025-02-02

    2023

  149. [156]

    Vladimir N. Vapnik. 1999. The Nature of Statistical Learning Theory . Springer New York

    1999

  150. [157]

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. Advances in neural information processing systems 30 (2017)

    2017

  151. [158]

    Fabrizio Ventola, Steven Braun, Zhongjie Yu, Martin Mundt, and Kristian Kersting. 2023. Probabilistic Circuits That Know What They Don’t Know. Uncertainty in Artificial Intelligence (UAI) (2023)

    2023

  152. [159]

    Richard von Mises. 1964. Mathematical theory of probability and statistics . Academic Press, New York, Chapter VIII.9.3

    1964

  153. [160]

    Ryan Webster, Julien Rabin, Loic Simon, and Frederic Jurie. 2023. On the De-duplication of LAION-2B. preprint arXiv:2303.12733 (2023)

    2023 arXiv

  154. [161]

    P. J. Werbos. 1982. Applications of advances in nonlinear sensitivity analysis. Systems Modeling and Optimization: Proc. IFIP (1982)

    1982

  155. [162]

    Rowan Zellers, Ari Holtzman, Yonatan Bisk, Ali Farhadi, and Yejin Choi. 2019. HellaSwag: Can a Machine Really Finish Your Sentence?. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics

    2019

  156. [163]

    Chiyuan Zhang, Samy Bengio, Moritz Hardt, Benjamin Recht, and Oriol Vinyals. 2017. Understanding deep learning requires rethinking generalization. In International Conference on Learning Representations

    2017

  157. [164]

    Weinberger, and Yoav Artzi

    Tianyi Zhang, Varsha Kishore, Felix Wu, Kilian Q. Weinberger, and Yoav Artzi. 2020. BERTScore: Evaluating Text Generation with BERT. arXiv:1904.09675 [cs.CL] https://arxiv.org/abs/1904.09675

    2020 arXiv

  158. [2015]

    AAAI Conference on Artificial Intelligence (AAAI) (2015)

    Never-Ending Learning. AAAI Conference on Artificial Intelligence (AAAI) (2015)

    2015

  159. [2019]

    Neural Information Processing Systems (NeurIPS) (2019)

    Can You Trust Your Model’s Uncertainty? Evaluating Predictive Uncertainty Under Dataset Shift. Neural Information Processing Systems (NeurIPS) (2019)

    2019

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.