monosov_lab

Account of Ilya Monosov; Bloomberg Distinguished Professor studying algorithms & circuits of biological and artificial intelligence at Johns Hopkins

[monosov_lab @MonosovLab]

Go Julia Pai, Fatih Sogukpinar et al.

Heavily revised work on astrocytes role in learning. Incredibly grateful for the team's hard work on this. Extended biology and theory now added!

https://t.co/1PAWq7EKFH

[monosov_lab @MonosovLab]

Congrats Arman Hatami and Romina Aalishah!

[Arman Hatami @Arman__Hatami]

Can a model transfer a hidden preference through random numbers?
Yes. We show subliminal learning occurs in vision-language models. Preferences and even wording pass from teacher to student via unrelated data when doing knowledge distillation.
WACV 2027
https://t.co/Tvamr5qTEd

[monosov_lab @MonosovLab]

The fact that visual bias is transferred among “AI” networks **without human intention** has deep relevance for application (eg, military decision making) and for thinking about **learning among/across biological social agents**!

Paper accepted at WACV.

https://t.co/mnWBjhMGs0

[monosov_lab @MonosovLab]

Our posters this SFN!!! Come see us and say hi! (img by fbl2.0)

[monosov_lab @MonosovLab]

Totally! It’s a constant struggle and refinement and hard work to communicate strongly but skeptically

[Julio Martinez @JulioMTNeuro]

Science communication is hard. Simplifying complex ideas often means losing important nuances. I struggle with this too. I wish I could say we’ll cure autism in five years, but that would feel dishonest. Science needs optimism, but also intellectual honesty.

[nature @Nature]

Nature research paper: The neural basis of flavour–nutrient learning

https://t.co/kuMC8HIoxi

[monosov_lab @MonosovLab]

Friends curious about ….psychedelics in medicine, neural circuits, and computational modelling of mental illness:

Come join us for a robust discussion. 

https://t.co/fJ26q6Uuxi

[Richard Sutton @RichardSSutton]

Brute force computation is not enough for intelligence. Clever new methods are needed. The science of AI is far from finished. IMHO

[UC Berkeley @UCBerkeley]

Neuroscientists from UC Berkeley, UCSF and Stanford ran parallel trials of psilocybin to see what behavioral effects were reproducible in mice. The findings challenge reported results and underscore the importance of reproducibility in psychedelic research and its potential therapeutic applications.
https://t.co/UvQIt3eEGi

[Annegret Falkner @Neurrriot]

Obviously a big news day, but here is some more! And maybe very topical since we have spent the last year as a country watching a bunch of knock-down, drag-out fights. https://t.co/9qFq8WAbHo

[Annegret Falkner @Neurrriot]

Maybe my favorite thing we have done yet! Estrogen oscillates network dynamics in the female PFC between rigid and flexible states. Seems like something basic/fundamental we should know already, but somehow we don't 🙃

[Dmitry Krotov @DimaKrotov]

Our new paper in Nature Communications reveals an interesting connection between generative diffusion models and Associative Memory.

It identifies 3 phases of generated samples:
👉 Memorization: reproducing training examples verbatim.
👉 Spurious states: discovering new patterns between stored memories, while maintaining a well-defined basin of attraction around those samples.
👉 Generalization: generating genuinely novel samples with flat energy landscape around them.

The surprising twist? What was traditionally considered a flaw in Associative Memory — spurious states — may actually be the first sign of generative capabilities.

It was fun to work on this project with @baophamhq, @gaboraya, Matteo Negri, @mj_zaki, and @LucaAmb, and I am happy to see it published!

Paper: https://t.co/kkXTePfIQ0

[monosov_lab @MonosovLab]

@DimaKrotov very cool Dima!

[monosov_lab @MonosovLab]

Dear @ForeignPolicy : does book actually use the term capitalism or does it rather concentrate on greed and corruption? I ask because the latter evils are products of communism. Thanks for explaining. I still love you guys

[Foreign Policy @ForeignPolicy]

A new book explores the capitalist basis of the Kremlin’s aggressive ideology. https://t.co/R85uOeshPd

[Jonathan Kipnis 🟦 @jonykipnis]

We are hiring 🥳🤩👍
We are looking for amazing colleagues from neuroimmune or glia world to join us!

[Alireza Modirshanechi @modirshanechi]

Super happy to announce that we have got an Oral at #NeurIPS! 🥳

Together with amazing @ben_eysenbach, Peter Dayan, & @cpilab, we proposed a notion of controllability to derive bounds linking unsupervised skill learning to downstream goal-reaching.

https://t.co/B5BiCwGcLl
1/8

[Nature Human Behaviour @NatureHumBehav]

Refining a new movement by practicing it upside down dlvr.​it/TW0ZfK

[monosov_lab @MonosovLab]

Agreed. Very cool

[Kepecs Lab @KepecsLab]

One of the coolest electrophysiology results ever...
rapidly flickering neural representations reflecting conscious perception

Huge congrats @doristsao & @HesseJanis!

[Jacob Prince @_jacobprince_]

Excited to share our new preprint:

Parametric neural control differentiates top neural network models of primate visual cortex
https://t.co/O9zTJxsAEJ

with an amazing team: @WangBinxu *, @thomas_fel_ , @akjags , @grez72, @talia_konkle , Parisa Vaziri, and the Livingstone Lab at HMS.

@KempnerInst, @PsychHarvard, @harvardmed (2/14)

[monosov_lab @MonosovLab]

Looking forward to reading!

[Neuron @NeuroCellPress]

Neuron continues its three-part series on neuroscience and AI with “Neuroscience and AI: Emerging Synergies”.
Explore how brains and AI are converging, from language models and scientific discovery to more brain-like AI systems.
The synergy deepens.
https://t.co/8oDK5wd6DD

[monosov_lab @MonosovLab]

Very interesting!!!

https://t.co/XhW38atfKC