Artificial Intelligence Research Centre, CitAI.
@citai.bsky.social
We are an interdisciplinary team at City, University of London specialised in the intersection between the development of novel AI techniques, AGI, and XAI with a keen interest in its legal, ethical and social impact. https://cit-ai.net/
The secret lies in groupoids, not groups! 🫥 😐 arxiv.org/abs/2609.13035
Groupoid-Based Internal State Representations for Reinforcement Learning with Local Symmetries
Symmetries play a central role in reducing the complexity of reinforcement learning problems, yet most existing approaches rely on fixed group actions or predefined state abstractions. Classical reinf...
arxiv.org
Indeed, women are "less salient" in a men dominated world. Just note that less salient stimuli extinguish more slowly.
Mel Andrews@bayesianboy.bsky.social · 4mo agoInterestingly, the models themselves offer pseudo-mathematical explanations for this effect, claiming that women authors are simply “less salient.”
The 'multiversity' ethic and the spirit of Capitalism, AI version. With apologies to M. Weber for a terrible paraphrasing. "When a student can produce a plausible term paper in twenty minutes using Claude Opus or Google Gemini,... 1/n www.persuasion.community/p/the-multiv...
The University As We Know It Is Finished
That’s a good thing.
persuasion.community
As studies show, those who don't understand 'AI' are more likely to adopt it, which explains why so many experts are opposed to it and why so many very AI-literate Canadians are refusing to adopt it. Maybe the gov needs to listen to experts and Canadians instead of the AI industry?
Tech Won’t Save Us@techwontsave.us · 4mo agoMark Carney wants Canadians to become more “literate” on AI. But is a lack of understanding really why people are against it? This week @hadrianmk.bsky.social joins @parismarx.com to dig into the Canadian government’s AI strategy and its many blind spots. Full ep: techwontsave.us/episode/333_...
“(…) it’s a fallacy to think that cognitive science can be automated. @olivia.science: ‘Doing theoretical work is very difficult, and sometimes researchers might fall for instant gratification (…) But if you ask AI to take over, you run into many risks ….” 1/2 www.ru.nl/en/research/...
Psychologists shouldn't replace thinking with AI | Radboud University
For some psychologists, it's becoming more common to use AI systems to replace human thinking in research. That's a very risky choice based on misconceptions, warn Iris van Rooij and Olivia Guest in a...
ru.nl
PALMS: A Computational Implementation for Pavlovian Associative Learning Models Simulation https://www.biorxiv.org/content/10.64898/2026.05.05.722899v1
Traditional reinforcement learning representations are procedurally efficient but lack structural richness. We enhanced the agent’s ontology and extended the conceptualisation of trajectories to provide a more nuanced view of task execution. 1/n www.sciencedirect.com/science/arti...
A representational framework for learning and encoding structurally enriched trajectories in complex agent environments
The ability of artificial intelligence agents to make optimal decisions and generalise them to different domains and tasks is compromised in complex s…
sciencedirect.com
🧵 New preprint led by @bingbrunton.bsky.social, @elliottabe.bsky.social, @lawrencehu.bsky.social We gave a worm brain control of a fly body and it walked What did we learn? Nothing, other than deep reinforcement learning is effective We call it the digital sphinx www.biorxiv.org/content/10.6...
I guess our paper hit a bit of a nerve 👀 @olivia.science, @marentierra.bsky.social et al. (2025). Against the Uncritical Adoption of 'AI' Technologies in Academia. Zenodo. doi.org/10.5281/zeno...
We then integrated SETs into a Structurally Enriched Trajectory Learning and Encoding (SETLE) architecture that incorporates a heterogeneous graph-based memory structure with multi-level relational dependencies, essential for task generalisation. www.sciencedirect.com/science/arti... 4/4
A representational framework for learning and encoding structurally enriched trajectories in complex agent environments
The ability of artificial intelligence agents to make optimal decisions and generalise them to different domains and tasks is compromised in complex s…
sciencedirect.com
Here we go doi.org/10.1016/j.ne... A representational framework for learning and encoding structurally enriched trajectories in complex agent environments @corinacatarau.bsky.social, @e-mondragon.bsky.social @ealonso.bsky.social Neural Networks, Available online 19 March 2026 Cheers!
Redirecting
doi.org
Esther Mondragón@e-mondragon.bsky.social · 7mo agoBecause there is AI beyond LLMs and generative AI arxiv.org/abs/2503.13194
Because there is AI beyond LLMs and generative AI arxiv.org/abs/2503.13194
A representational framework for learning and encoding structurally enriched trajectories in complex agent environments
The ability of artificial intelligence agents to make optimal decisions and generalise them to different domains and tasks is compromised in complex scenarios. One way to address this issue has focuse...
arxiv.org
Esther Mondragón@e-mondragon.bsky.social · 7mo agoNow accepted for publication in Neural Networks.
PALMS: Pavlovian Associative Learning Models Simulator
Simulations are an indispensable step in the cycle of theory development and refinement, helping researchers formulate precise definitions, generate models, and make accurate predictions. This paper i...
arxiv.org
Esther Mondragón@e-mondragon.bsky.social · 8mo agoApplications and code github.com/cal-r/PALMS-...
New preprint and simulator of associative learning attentional models. Have fun! 👁️ arxiv.org/abs/2602.07519 cal-r.org/index.php?id... #simulation #associative_learning #attention
PALMS: Pavlovian Associative Learning Models Simulator
Simulations are an indispensable step in the cycle of theory development and refinement, helping researchers formulate precise definitions, generate models, and make accurate predictions. This paper i...
arxiv.org
Also, including me.
Chris Stokel-Walker@stokel.bsky.social · 8mo agoHow are governments integrating AI into their decision-making and legislative processes? I spoke to those on the front line, including the Estonian prime minister, for my latest for @techpolicypress.bsky.social techpolicy.press/governments-...
We are demanding that EU co-legislators reject attempts in the AI Omnibus to remove a key transparency safeguard from the AI Act. We cannot open a loophole that would let providers exempt themselves from the AI Act’s high risk requirements with no transparency www.accessnow.org/press-releas...
Access Now - A call to EU legislators: protect rights and reject the call to delete transparency safeguard in AI Act
We, the undersigned organisations and individuals, urge you in the strongest possible terms to reject the deletion of the Article 49(2) transparency safeguard for high-risk AI systems that is proposed...
accessnow.org
The term ‘Artificial Intelligence’ (AI) means many things to many people (see Table 1). (...) One meaning of ‘AI’ that seems often forgotten these days is one that played a crucial role in the birth of cognitive science as an interdiscipline in the 1970s and ’80s." 2/n
This is NOT what AI was about "What can researchers do if they suspect that their manuscripts have been peer reviewed using artificial intelligence (AI)?"
Shannon Vallor@shannonvallor.bsky.social · 10mo agoMany, many unpaid labour-hours will be required for the ICLR authors and organizers to sort this mess out, and the conference will be worse for it. A perfect case study on why AI productivity gains are largely a mirage.
This -> "We have deprived the youth of the joys of discovery, the thrill to feel yourself close to an answer, and the wonderful, rewarding feeling of a moment of lucidity after hard work."
Esther Mondragón@e-mondragon.bsky.social · 10mo agoPerhaps of interest to some of you. "Quick societal adoption of tools often reflects a demand driven by business interests or culture. Reflecting on generative AI, especially Large Language Models, I believe there is a strong lifestyle generation component that nurtures.. cal-r.org/mondragon/S8...
Perhaps of interest to some of you. "Quick societal adoption of tools often reflects a demand driven by business interests or culture. Reflecting on generative AI, especially Large Language Models, I believe there is a strong lifestyle generation component that nurtures.. cal-r.org/mondragon/S8...
Scripto_Asinine or sound minds?
cal-r.org
Ablation study and resulting optimal architecture that considerably improves recent research on CNN-Hebbian learning integration with competition mechanisms. with @julian-jn.bsky.social www.sciencedirect.com/science/arti...
Advancing the Biological Plausibility and Efficacy of Hebbian Convolutional Neural Networks
The research presented in this paper advances the integration of Hebbian learning into Convolutional Neural Networks (CNNs) for image processing, syst…
sciencedirect.com
Surprise, surprise. Here we go. “The European Commission is proposing a pause to parts of its landmark artificial intelligence laws amid intense pressure from Big Tech companies and the US government.” www.ft.com/content/af6c...
EU set to water down landmark AI act after Big Tech pressure
Commission proposes pauses to provisions in digital rule book
ft.com
✨ This is wonderful 🎬 🍿 Librebel on Youtube reads out our position paper: Guest, O., Suarez, … & van Rooij, I. (2025). Against the Uncritical Adoption of 'AI' Technologies in Academia. Zenodo. doi.org/10.5281/zeno... www.youtube.com/watch?v=cJNO... @olivia.science @marentierra.bsky.social
Against the Uncritical Adoption of AI Technologies in Academia by Guest et. al. (2025)
YouTube video by librebel
youtube.com
“Ultimately, the collective strategy of AI companies threatens to deskill precisely those people who are essential for society to function(…) automation of knowledge and culture by private companies is a worrying prospect – conjuring dystopian and outright fascistic scenarios.” — @olivia.science
AI Is Hollowing Out Higher Education
Olivia Guest & Iris van Rooij urge teachers and scholars to reject tools that commodify learning, deskill students, and promote illiteracy.
project-syndicate.org
A quick summary for busy people: We propose a mathematical framework for learning representations by extracting the algebra of transformations of worlds from the agent's perspective. 1/n www.sciencedirect.com/science/arti...
Algebras of actions in an agent's representations of the world
Learning efficient representations allows robust processing of data, data that can then be generalised across different tasks and domains, and it is t…
sciencedirect.com
Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia: doi.org/10.5281/zeno... We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical thinking, expertise, academic freedom, & scientific integrity. 1/n
I am pleased to announce that this paper has been accepted for publication in Artificial Intelligence (AIJ)! 😊 arxiv.org/abs/2310.01536
Algebras of actions in an agent's representations of the world
In this paper, we propose a framework to extract the algebra of the transformations of worlds from the perspective of an agent. As a starting point, we use our framework to reproduce the symmetry-base...
arxiv.org
Esther Mondragón@e-mondragon.bsky.social · last yr.New version available. The readability of the paper has been greatly improved. Enjoy! arxiv.org/abs/2310.01536
New version available. The readability of the paper has been greatly improved. Enjoy! arxiv.org/abs/2310.01536
Algebras of actions in an agent's representations of the world
In this paper, we propose a framework to extract the algebra of the transformations of worlds from the perspective of an agent. As a starting point, we use our framework to reproduce the symmetry-base...
arxiv.org
