Endorse, and I'd also add that it's nice to mention a few things you liked about the paper at the start of the review.

My reviewing style has changed over time. Rather than litigate every little thing, and pushing my own ideas, I focus only on 2 things: (1) Are the claims interesting/important? (2) Does the evidence support the claims? Most of my reviews these days are short and focused.

Talks from #SNUFA 2025 are now available on YouTube: youtube.com/playlist?lis... 🤖🧠🧪

"studies have found little evidence for uniformly increased sensory precision in autism, but instead point to aberrant context-sensitive modulation of prediction errors"? which explains "why autism is associated with both hyper- & hyporeactivity to sensory input"? www.cell.com/trends/cogni... review

Watson was a racist who, "near the end of his life, faced condemnation and professional censure for offensive remarks, including saying Black people are less intelligent than white people"

I saw Jim Watson speak in the flesh only once, in 2013. During the talk he mentioned the "historic curse of the Irish, which is not alcohol, it’s not stupidity. But it’s ignorance." www.sandiegouniontribune.com/2013/03/21/w...

Delighted to announce with my colleagues Prof Yvonne Daly, Dublin City Uni; Prof Dave Walsh Walsh, De Montfort Uni; Prof Bennett Kleinberg, Tilburg Uni, we have secured an ERC Synergy grant entitled JUSTICE (or ‘Joining Unique Strategies Together For Interrogative Coercion Elimination).

The world’s first trillionaire initiated a move that has left more than half a million people dead, most of whom are children.

Personally I have never used an LLM to: - review a paper - review a grant - mark student assessments These are tasks I do not pass to LLMs because I am getting paid to do them and because I was asked to provide my individual professional judgement on their quality

LLMs are already reviewing science... just often not declared

My prediction is that LLM peer review will slow down science. It will do this for precisely the same reasons that contemporary peer review does and some extra ones. Start by reading @hansonmark.bsky.social thread below, then read on. 🧵

The advances we've made in statistics, experimental study design, and causal inference over the past century are remarkably useful for understanding our world. But there is never been a push to make people use them like we are seeing with generative AI. Perhaps take a moment to consider why.

Getting nervous for the talk I'm about to give at a workshop about "using AI to drive impact" which features slides such as these.

We must change this mindset that the essential and most critical work in a field is always the first one. That distorted view has caused a lot of problems in psychological science already. The first work is just that—the start. Every field should be evaluated based on the complete body of evidence.

Reactions to the "debunking" of the original 1956 cognitive dissonance work baffles me. A whole area doesn't die just bec the first work is faulty. That's novelty bias & theory ownership in psych talking. Mendelian genetics did not die when Mendel's original results were found too good to be true.

My favorite comment on the FT story

how much of this concession is just shareholder fear that "if Musk walks away our stock value will fall"?

Elon Musk is boosting the British right - and this shows how

not the first and almost certainly not the last time I am asking ChatGPT: what's the difference between a research aim and an objective

2/n Scientific heroes from the past had v different experiences / opportunities. His wife "..ran their households, hosted exuberant parties and raised their children. In short, she provided an environment in which Crick had the luxury of devoting all of his energy to his intellectual life." and..

🌟Prof @lizcoulthard.bsky.social 🌟discussing blood tests for dementia www.bbc.co.uk/news/article...

Many of our big insights into brain function come from trying to mimic it, writes @timothyoleary.bsky.social. This lesson should guide how we organize research programs. www.thetransmitter.org/systems-neur... #neuroskyence

Introducing CorText: a framework that fuses brain data directly into a large language model, allowing for interactive neural readout using natural language. tl;dr: you can now chat with a brain scan 🧠💬 1/n

Europe's MSCA postdoctoral fellowships are now attracting many researchers who would have gone to the US. So the payline has dropped below 10% (~1,600 fellowships to 17k applicants). Super tough situation. One more reason to double the next Horizon budget, as proposed by the EU Commission.

How does the brain find its way in realistic environments? 🧠 Using deep RL and neural data, we show that hippocampal-like networks support navigation, learning, and generalisation in partially observable environments—mirroring real animal behaviour. Now out: www.nature.com/articles/s41... #neuroAI

I’ve spent many years explaining/defending biorxiv’s “no reviews” policy. The logic was always that opening Word and opining is a far lower barrier than doing actual research, so noise’d be >> signal and we didn’t want to make subjective quality judgements. LLMs mean it makes even more sense 1/2

An MP needs to read out this quote in the House of Commons and ask the government why they are still on twitter cc @columeastwood.ie

Elon Musk, "These lovely small towns in England, Scotland and Ireland, they've been living their lives quietly. They're like hobbits" "And so one day, 1,000 people show up in your village of 500 and start raping the kids" "This has now happened, God knows how many times in Britain"

I was driving my kids to school one day and they were fighting in the back seat, so I said "ok, I want you to say 3 things you like about the other person". Bit of silence, then one of them said... "can we start with don't likes?"

Reminder everyone that #SNUFA is next week! Make sure to register in advance to get emailed the link to join.

It's only Donegal tweed if it's from the Tír Chonaill region of Ireland. Otherwise it's just sparkling wool 😉 Good article on the fight to get Donegal tweed a protected status www.ft.com/content/54f5...

Excited to share the latest from the lab on the molecular mechanisms of silencing and un-silencing a hippocampal engram ensemble. Was fun using a variant of the EPSILON technique developed by Adam Cohen's lab. (1/2) www.sciencedirect.com/science/arti...

Thrilled that our paper is out today in Nature! www.nature.com/articles/s4...

My co-authors have yet to move to Bluesky, so I'm pleased to announce our latest work has just been published in @nature.com Neuroscience. Amazing work led by Junheng Li, revealing that falling asleep follows a predictable bifurcation pattern #neuroskyence #sleep www.nature.com/articles/s41...

Just monstrously bad

Assuming that everyone learns the way you do, and then structuring your pedagogy on that premise, it's one of the worst things you can do when working with any students

When #HHMIJanelia released the #Drosophila hemibrain in 2020, rendering all cell types at once with full shading/shadows was too hard, so the video showed them region by region. Technology has since advanced, and I went back and rendered them all together. I like how it shows the internal structure.

The question isn’t "why does Signal use AWS?" It’s to look at the infrastructural requirements of any global, real-time, mass comms platform and ask how it is that we got to a place where there’s no realistic alternative to AWS and the other hyperscalers. 3/

> “Universities are not tech companies,” Olivia Guest and Iris van Rooij, two computational cognitive scientists at Radboud University in the Netherlands[.] “Our role is to foster critical thinking,” the researchers said, “not to follow industry trends uncritically.”

Not to be that person but from the original Box (1976)

New preprint! What happens in the brain when people offload memories into external reminders? Using fMRI decoding, we found that the corresponding neural trace fades until it becomes statistically absent. osf.io/preprints/ps... 🧵...

The home of the O'Donovan family is being used as a polling station in Coolmoyne, Co Tipperary. Around 250 people are eligible to vote at the house and after they cast their ballot, they will be treated to a cup of tea, a slice of apple or rhubarb tart, or a scone | More: rte.ie/b/1540324

We also support and ship solvers from SUNDIALS blogs.mathworks.com/matlab/2024/... Which of those 50 solvers are useful to you and why?

George Box famously said "all models are wrong, some are useful", but what he forgot to add was that usefulness doesn't just depend on the model. A model is useful *only with respect to a given target problem*

Spiking NN fans - the #SNUFA workshop (Nov 5-6) agenda is finalised and online now. Make sure to register (free) soon. (Note you can register for either day and come to both.) Agenda: snufa.net/2025/ Registration: www.eventbrite.co.uk/e/snufa-2025... Thanks to all who voted on abstracts! 🤖🧠🧪

Join the team at Janelia to use #theory, computation, and modeling to tackle bold questions in #biology. Collaborate across scientific disciplines to drive discovery. Apply by Nov. 3 @ https://janelia.link/theoryfellowprogram #computationalbiology #sciencecareers

Ireland has a presidential election this Friday (largely ceremonial position), 2 candidates and the left-leaning independent one Catherine Connolly is well ahead in the polls. Presumably because of this she is being hit hard this week by newspapers but now disturbingly also deepfake disinformation

odd framing in this (otherwise v good) BBC article on Lough Neagh pollution. Early on quotes Mellor listing 6 sources with agriculture named 3rd Later describes report from ABFI (Mellor's employer) that says agriculture responsible for 60% of nutrient load in the lake www.bbc.co.uk/news/article...

For this reason I prefer the UK PhD model of 3-4 years with fairly hard time limit. Do you learn less and achieve less than in a US PhD? Probably. Do you still learn *enough* to call yourself an expert in a topic? Usually, yes. Let people move on with their lives