Minh-Huy Mai-Duc

Minh-Huy Mai-Duc

@sabertoaster.bsky.social

Undergrad @ VNUHCM University of Science. Interested in doing NeuroAI research. Happy to discuss and for further reaching out:) ---- 2026 will be my year!

55 followers204 following44 posts

Minh-Huy Mai-Duc
@sabertoaster.bsky.social

Super entertaining and informative(^^) , especially on Diogo's take on common hatred towards LLMs > [...] we don't hate technology, we just hate when it's bad for us typesafe.ai/blog/ai-too-...

typesafe.aiAI: too good to be true, too bad to be useful | TypeSafe AI - TypeSafe AI BlogTypeSafe AI is an AI lab building machine-native intelligence infrastructure for automation, designed to make decisions within software. Try our first System One Model, Jev, in early access.

Minh-Huy Mai-Duc reposted

allen institute
@alleninstitute.org

AI neurons run on simple math. Real neurons are far weirder. A new open dataset fitted 1,000+ PyTorch-ready neuron models to real mouse and human cells. Does brain-like diversity make AI better? It's still an open question, but now it's testable. https://bit.ly/3WfGAwR

Minh-Huy Mai-Duc
@sabertoaster.bsky.social

New preprint out from my time at Omnibio, in joint work with Robert Clarke (University of Bath). We introduced Weighted Stochastic Block Model mechanism to Echo State Network and investigated its coupled dynamics in Reinforcement Learning environments. arxiv.org/abs/2610.15380

arxiv.orgAnalyzing the Dynamics of WSBM-based Echo State Networks using InputDSABridging the structural gap between biological brains and artificial neural networks presents a pathway to altered computational dynamics. This paper introduces a Reservoir Computing framework that re...
Minh-Huy Mai-Duc
@sabertoaster.bsky.social

The preliminary results showed correlation between high-low performing agents and the disparities in its (now decoupled thanks to InputDSA) intrinsic dynamics. This is an initial exploration, and we are actively extending the framework. Feedback and discussions are welcome.

Minh-Huy Mai-Duc reposted

Nic 💀 "skull" 💀 Dav
@nicholdav.bsky.social

Writing something up, and reading this + re-reading Mensh & @kordinglab.bsky.social's: journals.plos.org/ploscompbiol... The new paper has steps to follow, helps my ADD brain. The older paper helps me keep my eyes on the prize. I feel like many researchers need these, like, now more than ever

journals.plos.orgTen simple rules for structuring papers
John M. Drake@jdrakephd.bsky.social

Newly expanded version of my guide to scientific writing -- known as the “15 steps” -- published in PLOS Computational Biology. Special thanks to Éric Marty for creating a fantastic visualization. Check it out: journals.plos.org/ploscompbiol... #ScientificWriting #PLOSComputationalBiology

Minh-Huy Mai-Duc reposted

Olivia Guest · Ολίβια Γκεστ
@olivia.science

Replying to @olivia.science

Which is also why I made this figure for this relevant paper (tiny btw only like 2 pages). Guest, O., Blokpoel, M.& van Rooij, I. (2026). What the Func? Multiple Realizability Need not be Vague. Zenodo. doi.org/10.5281/zeno... 2/2

Iris van Rooij 💭@irisvanrooij.bsky.social

We propose 4 different notions of ‘function’ and formalise each. 1. Functional capacity 2. Functional theory 3. Functional goal 4. Functional role --- Preprint: doi.org/10.5281/zeno... 4/🧵

Minh-Huy Mai-Duc
@sabertoaster.bsky.social

I'm still feeling like I've not committed fully to NeuroAI from a computational neuroimaging (CN) perspective. My thought is, I've never saw any direct implications in CN to NAI (therefore I'm lacking motivation), so if someone has any views / regards for this Id be very appreciated ^^

Minh-Huy Mai-Duc reposted

Blake Richards
@tyrellturing.bsky.social

1/2) Please let's all remember that GWT is a theoy of access consciousness (what is available for verbal reasoning) and not phenomenal consciousness (subjective experience): philpapers.org/rec/BLOOAC Obviously, in this study access consciousness is literally what they test in the LLMs. #NeuroAI 🧪

philpapers.orgphilpapers.org
Anthropic {bot}@anthropicai.xmirror.bot

New Anthropic research: A global workspace in language models. Of everything happening in your brain right now, only a tiny fraction is consciously accessible—thoughts you can describe, hold in mind, and reason with. (1/2)

Minh-Huy Mai-Duc
@sabertoaster.bsky.social

There's a project named "Synthetic pleasure: Can a robot have an orgasm?" :)

Neuromatch@neuromatch.bsky.social

🎉 Introducing the 2026 Neuromatch AI Sentience Scholars cohort! Learn more about them, the projects they will be working on, and their mentors! ⬇️ neuromatch.io/ai-sentience... We're just kicking off this 6 month program and already incredibly proud! #AISentience #AIResearch #ConsciousnessStudies

Minh-Huy Mai-Duc reposted

Dan Levenstein
@dlevenstein.bsky.social

“under any single goal, the human connectome is suboptimal. Instead, its organization reflects multi-objective trade-offs navigated over evolution and development under biological and environmental constraints” Not just connectome…

CLaE@claeneuro.bsky.social

Trends in Cognitive Sciences Embracing the suboptimal organization of the human brain. www.cell.com/trends/cogni...

cell.comEmbracing the suboptimal organization of the human brainHuman brain architecture is guiding brain-inspired artificial intelligence (AI) and has been treated as an optimal template, whose deviations could mark different psychiatric and neurological conditio...

Minh-Huy Mai-Duc reposted

Thomas Serre
@thomasserre.bsky.social

Published @cp-trendscognsci.bsky.social with @drewlinsley.bsky.social & @tonyfeng.bsky.social: As vision models scale to human/superhuman accuracy, they’re becoming worse models of primate vision—benchmark engineering isn’t neuroscience. @carneyinstitute.bsky.social @browncopsy.bsky.social

cell.comBetter artificial intelligence does not mean better models of biologyDeep neural networks (DNNs) once showed increasing alignment with primate perception as they improved on vision benchmarks, raising hopes that advances in artificial intelligence (AI) would naturally ...

Minh-Huy Mai-Duc reposted

Mackenzie Weygandt Mathis
@trackingactions.bsky.social

Interested in the latest advances in neuroscience (neural dynamics and internal models) and how they can be leveraged to build smarter, adaptive AI? ➡️ My first real solo piece 🖤🫶 @natneuro.nature.com rdcu.be/eWVmA

rdcu.beLeveraging insights from neuroscience to build adaptive artificial intelligenceNature Neuroscience - Adaptive intelligence envisions AI that, like animals, learns online, generalizes and adapts quickly. This Perspective reviews biological foundations, progress in AI and...

You are up to date.