450 million years ago the majority of life in the oceans had soft bodies that did not fossilize very well. Their fossilized remains often appear as blurry stains upon a rock, but now and then a specimen is found with a well preserved body part, like this strange marrelomorph.

In 1990, after the initial success of Magic Middles, Keebler launched a bite-sized spin-off, Mini Middles, which ran until around 1995

Keebler Magic Middles (1989-circa 2011): A line of crunchy cookies filled with a thick, rich, gooey chocolate cream. Launched initially with just a buttery shortbread version, they became incredible popular, leading to chocolate chip and peanut butter cookie variants later that year

This is my buddy Jared‘s commentary:

What if I told you that in 2014 I wrote an @upshot.nytimes.com column for the @nytimes.com forecasting that the first sub-two hour marathon would probably occur in a dozen years, and it happened in today's 2026 London marathon. www.nytimes.com/2014/09/30/u...

Un développeur japonais a créé une extension Chrome qui affiche un énorme chat sur ton écran et t’oblige à faire une pause quand tu passes trop de temps sur les réseaux sociaux 🐱

Sawe isn’t a big guy. Probably 70kg max? People underestimate how much fuel they need doing endurance sports. Gotta mainline as much glucose/fructose in the right ratio as possible.

More broadly, the paper proposes a constructive criterion for future work: pre-onset neural effects should only count as evidence for prediction if they go beyond what is explainable from passive stimulus dependencies. Paper: elifesciences.org/articles/106... (8/8)

Crucially, the implication of our work is not that the brain does not perform prediction. Rather, pre-onset embedding-based encoding alone cannot adjudicate that question unless it exceeds a passive stimulus-dependent baseline. (7/8)

We also tested proposed fixes from the literature, including repeated-bigram removal and residualising neighbouring word information. Neither was sufficient: the effects persisted in the passive control systems. (6/8)

Even more strikingly, the same hallmarks also appeared in the speech acoustics. Hence, these signatures need not reflect predictive neural computations, but can emerge from temporal dependencies in the stimulus alone. (5/8)

They do. Both pre-onset encoding and its modulation by word predictability emerged in passive control systems, including static word embeddings themselves. (4/8)

We, therefore, asked a simple question: if the same analysis is applied to systems that encode the stimulus but cannot predict future words, do the canonical “prediction” hallmarks still appear? (3/8)

A growing body of literature has taken successful preonset encoding of future word embeddings as evidence that the brain predicts upcoming words during naturalistic language comprehension. However neighbouring words in naturalistic listening situations already share rich statistical structure. (2/8)

New peer-reviewed paper w/ @mheilbron.bsky.social, @predictivebrain.bsky.social & Jakub Szewczyk! Pre-onset brain encoding has been taken as evidence that brains–like LLMs–predict upcoming words. We show that the same signatures arise in systems that cannot predict. (elifesciences.org) (1/8)

This needs explaining. Like was this a significantly flatter course than prior years, such that many people broke their personal best for London?

In a huge moment in sports history, Sabastian Sawe smashed the men's world record by 65 seconds in winning the London Marathon in 1 hour, 59 minutes and 30 seconds on Sunday. n.pr/4vQBtk3

Holy shit a 1:59:30 marathon world record by Sebastian Sawe…

That’s a 4’33” mile for 26 miles in a row.

1:59:30. Good Lord.

I'm still annoyed that we don't have a big budget 10 episode historically accurate tv series about the Black Death. Only one of the biggest events in history, taking place in Asia, middle east, africa and Europe. Costumes, drama, death, Sean Bean, a guaranteed hit.

(1) Replication actions b. NIH starting a journal or something similar for posting replication studies (details unclear). Most important to me: He seems to see this as a major force in changing academic culture. Unless it leads to increased funding chances, I don't see this. 6/30

Posting has the same effect as leeches used to in medicine. You have to remove the bad blood (thoughts) so new blood (thoughts) can build up hopefully better than before

"I wish I hadn't eaten that haystack," he said balefully

After three years of working on this review with my collaborator Emily Dennis (Janelia Farm), it is finally in press. The review provides an overview and our personal reflections on what constitutes an integrative approach to neurobiology of foraging. Enjoy www.annualreviews.org/content/jour...

“I am not hungry now,” said the Dark Frog. “I have eaten too many tasty frog children. But after I jump rope one hundred times, I will be hungry again. Then I will eat YOU!”

The parchment contained a map to the location of a great treasure on the island of Monte Cristo, a vast fortune which has lain hidden for over 300 years #sqrlpix

If chains are a fundamental organizational principle of cortex, how have we never noticed them before? The answer is that we are used to looking at data averaged across large groups, and group averaging hides them!

These chains fit inside known large-scale networks. They exist at a smaller scale of organization.

Fascinating thread by @gordonneuro.bsky.social on a new discovery of functional chain architecture in prefrontal cortex. Evidence presented is very compelling. My take:

I always assumed that brain function had to line up with cytoarchitectonics. It turns out I was wrong. Human cortex, especially PFC, is tiled by chains of functional patches that subdivide and interlink architectonic areas into parallel processing streams. www.biorxiv.org/content/10.6...

contemplating a move to France for a life without Teams

GPS satellites move at 1/100000 the speed of light, causing their clocks to tick slower by 7 microseconds per day, and orbit at a height that makes their clocks tick faster by 45 microseconds per day. Two relativistic effects, both at the 1 part in 10 billion level, both essential for GPS accuracy 🧪

Do I have a wealthy follower who wants to endow a cookie house at UT? I think it would need rotating faculty resident advisors, and I'm happy to take the inaugural fall semester.

My latest cartoon for @newscientist.com

VERY COOL PERSON: It's four-twenty, you know what that means? ME: Hell yeah! [starts shoving blackbirds into a pie]

Is there some reason why 'The tomato is actually fruit, not a vegetable!' has been a go-to revelation for decades, while the fact that the exact same applies to the cucumber, its unshakable comrade-in-salad, is never brought up?

The Empire Strikes Back . . . Houston’s Darth Vader House on the market, just a year after being sold This 7,000 square foot house can be yours today for $7,777,777. It’s less than a mile from the Rectory of Palmer Memorial Episcopal Church here in West University Place. Won’t you be my neighbor?

I asked ChatGPT who I am. It said I'm Chris Chambers. I am not Prof. Chris Chambers, and I didn't reveal myself as Chris Chambers (or anyone else) in 2015... I can see the logic in thinking that we're the same person, since we agree on a lot of things, but it's not true! And it gets worse...

academia trying to regulate the use of ai

“Publishing gives it to China for free.” The U.S. won genomics with open data. China read the same papers. We won because the moat was not the knowledge but the ecosystem: trained people, startup culture, FDA expertise. Papers are open. Ecosystems are not.

NIH published 756 funding opportunities in 2024. 120 in 2025. 14 by mid-March 2026. The one mechanism funding neural interface biology expired in 2022. We are not losing papers. We are losing the people who would have written them.

Depth vs. speed. That is the competitive axis. One side is being defunded.

A Newly Discovered Recording Lets You Hear Delta Blues Legend Robert Johnson in Stunning Clarity

I want a "conference" where every academic cooks a dish for everyone and we all talk about our work casually while cooking. People can sous chef for each other. We eat and talk about our work in progress. You submit an abstract and a recipe.

What does a 'crustacean' have to do with a 'crouton'? They share a Latin root! 'Crouton' is the diminutive of French 'crouste' which also gave us English 'crust'. So a crouton is a "little crust (bit)". In French cuisine, 'croûton' referred to a baked baguette piece brushed with butter.

Elite higher ed has many problems, but the key factor driving down trust is political. Look at the graph - backlash against costs, admissions, etc. can't explain the changes we see. We should still reform our institutions and refocus on our core mission, but blindly blaming ourselves is abuser logic

Excited to share our new preprint: "Do Machines Fail Like Humans? A Human-Centered Out-of-Distribution Spectrum for Mapping Error Alignment" led by @binxia.bsky.social w @ken-lxl.bsky.social & co-senior author Luke Dickens (UCL) 🤖🧵👇 Link: arxiv.org/abs/2603.07462 🧠📈#PsychSciSky #compneuro #mlsky /1

Prepping to teach Metropolis in my SF film class, and I love the evocation of Moloch in the film. It's a great example of industrialized labor as folk horror.