Federico Adolfi

I’m interrupting this transmission to let you know that as I was walking around in a poster session the other day I realized I cannot take seriously any work with “Lotka-Volterra” on it and it’s all due to some joke you probably made 4 years ago

This reminds me of that famous paper ‘Neuroscience needs no stinking behavior’

Turbulence

I just discovered (the hard way) that having a fruit salad before drinking tonic water makes the latter taste like airplane toilet

@laurennross.bsky.social is giving an important reminder of the importance of philosophy of science in cognitive science at the #cogsci2025 keynote, which I really appreciate in the current moment.

Me: hi sorry I think I left my keycard in my room

Front desk guy: no problem sir what’s your room number

Me: I think I forgot my room number

Guy: can I have some id

Me: look we can do this all day

Me: hi sorry I think I left my keycard in my room

Front desk guy: no problem sir what’s your room number

Me: I think I forgot my room number

Guy: can I have some id

Me: look we can do this all day

If I may level up to the galaxy brain variant:

Just trotting out the aphorism “all models are wrong but some are useful” doesn’t make your…thing… a model

Brain Surfaces of 70 primate species

1
To predict the behaviour of a primate, would you rather base your guess on a closely related species or one with a similar brain shape? We looked at brains & behaviours of 70 species, you’ll be surprised!

🧵Thread on our new preprint with @r3rt0.bsky.social , doi.org/10.1101/2025...

Kaiserslautern, Germany

📣 Life update: Thrilled to announce that I’ll be starting as faculty at the Max Planck Institute for Software Systems this Fall!

I’ll be recruiting PhD students in the upcoming cycle, as well as research interns throughout the year: lasharavichander.github.io/contact.html

Excited to announce the first workshop on CogInterp: Interpreting Cognition in Deep Learning Models @ NeurIPS 2025! 📣

How can we interpret the algorithms and representations underlying complex behavior in deep learning models?

🌐 coginterp.github.io/neurips2025/

1/4

Psychologist: does vision rely on texture or shape?

Cognitive neuroscientist: I’ll check which of the two this brain region cares about

Next gen naturalistic comp cog NeuroAI person: shh guys enough with the dichotomous testing, I’m busy testing if it’s the training data or the objective function

Psychologist: does vision rely on texture or shape?

Cognitive neuroscientist: I’ll check which of the two this brain region cares about

Next gen naturalistic comp cog NeuroAI person: shh guys enough with the dichotomous testing, I’m busy testing if it’s the training data or the objective function

And cynicism! Don’t forget cynicism

Welcome to July, the last month of the year we get “sorry I meant July of course not June” emails

Important conversation and insights here. But the most important thing we should learn from this thread is how effectively I have so far dodged the monumental nerd snipe that this is for me

Ok I fully sympathize and share this experience but also:

bsky.app/profile/fede...

ExplainableAI has long frustrated me by lacking a clear theory of what an explanation should do. Improve use of a model for what? How? Given a task what's max effect explanation could have? It's complicated bc most methods are functions of features & prediction but not true state being predicted 1/

Very cool, look forward to reading it! You might be interested in our recent ICLR paper, which approaches circuit discovery in mechanistic interpretability in a similar spirit re formalization and attention to the affordances of explanations (through diff methods, tho)
openreview.net/forum?id=Qog...

I have a lot of concerns about the Binz et al. Centaur paper, but I'll start this thread with what I like: Matching human behavioral data is critical when modeling human cognition. Trying to match neural data is interesting and possibly informative, but not sufficient for modeling the human mind. 1/

The scene from Severance where Mr. Milchick sits in front of a mirror and practices saying things using shorter phrases and simpler words. The caption reads: "my scholarly pursuits deserve your praise and wealth...", "my work is mysterious and interesting...", "I did a thing...", "Fund me".

Me writing a grant

Me: so we made an honest effort and wrote a research statement. Now we can rel—

Brain:

Me: —what?

Brain:

Me: don’t be dramat—

Brain: I hate it and wanna set it on fire

Sometimes I feel guilty that I’m posting my silly jokes here and then I see others posting like they are crossing the worst reverse inference fallacies from cognitive neuroscience with the most lazy takes on LLMs and I’m like you know what it’s probably ok people need a break from this

Binz et al. (in press, Nature) developed an LLM called Centaur that better predicts human responses in 159 of 160 behavioural experiments compared to existing cognitive models. See: arxiv.org/abs/2410.20268

Sometimes I feel guilty that I’m posting my silly jokes here and then I see others posting like they are crossing the worst reverse inference fallacies from cognitive neuroscience with the most lazy takes on LLMs and I’m like you know what it’s probably ok people need a break from this

Me: so we made an honest effort and wrote a research statement. Now we can rel—

Brain:

Me: —what?

Brain:

Me: don’t be dramat—

Brain: I hate it and wanna set it on fire

The scene from Severance where Mr. Milchick sits in front of a mirror and practices saying things using shorter phrases and simpler words. The caption reads: "my scholarly pursuits deserve your praise and wealth...", "my work is mysterious and interesting...", "I did a thing...", "Fund me".

Me writing a grant

Me: put that down already, we’re working on project B today

Brain: *staring at shiny project A* why not, after all, why shouldn’t I keep it

Me: ffs leave it

Brain: my preci—

Me: don’t

I don’t believe my surname has any nice resonances (especially not in Germany 😬) but you have to admit Gandolfi is a great candidate for Italian Gandalf

A handmade brown suede notebook

Brown suede notebook cross stitched on the side

I’m sorry for any psychic damage this may cause but in fairness it’s your fault for being the professor of industrial design that asked the question and a bit my fault for winging it. It’s supposed to be a notebook