Eoin Travers

It'ssssss PREPRINT THURSDAY! (Sadly not a thing). Approximately an era ago, I spent a *lot* of time discussing the assumptions that underpin the field of #ComputationalPskychiatry with the brilliant @jonroiser.bsky.social and @olijrobinson.bsky.social osf.io/preprints/ps...

I think deep linear networks are an example of this, where you have a deep model with just the capacity of regular linear regression.

Bullshitting Engines?

To be fair, if you're doing analytic philosophy, "bullshit engine" reads as an engine that is bullshit, not an engine that engages in the communicative act of bullshitting, because engines don't communicate or have concerns, they set things in motion.

Grid Cells for Conceptual Spaces? “Grid cells” encode an animal’s location and direction of movement in 2D physical environments via regularly repeating receptive fields. Constantinescu et al. (2016) report the first evidence of grid ...

Sorry, broke the second link: www.cell.com/neuron/fullt...

Organizing Conceptual Knowledge in Humans with a Grid-like Code It has been hypothesized that the brain organizes concepts into a mental map, allowing conceptual relationships to be navigated in a similar fashion to space. Grid cells use a hexagonally-symmetric co...

Yup: pmc.ncbi.nlm.nih.gov/articles/PMC... (commentary: www.cell.com/neuron/fullt.... Lots of later work on the same idea since then as well.

Today, we’re announcing the preview release of ty, an extremely fast type checker and language server for Python, written in Rust.

In early testing, it's 10x, 50x, even 100x faster than existing type checkers. (We've seen >600x speed-ups over Mypy in some real-world projects.)

It's in the opposite direction though, once your participants know how you want them to behave, they're pretty likely to behave that way.
en.m.wikipedia.org/wiki/Demand_...

I think it's a choose-your-own-null adventure kind of thing, like you're worried it might be. I can see the (bad) argument, if you allow lots of parameters to vary either side of the dotted line, any differences prove that the line is important?

Um.
statmodeling.stat.columbia.edu/2021/11/21/s...

Why High-Order Polynomials Should Not Be Used in Regression Discontinuity Designs It is common in regression discontinuity analysis to control for third, fourth, or higher-degree polynomials of the forcing variable. There appears to be a perception that such methods are theoreti...

If you think different slopes are bad, you should see www.tandfonline.com/doi/abs/10.1... (and related posts: statmodeling.stat.columbia.edu?s=regression...). Polynomial madness.

I'm on the job market, and like many, I hate LinkedIn. So I've done what any product-oriented data scientist would: built my own workflow. With browser automation/scraping, data pipelines, structured LLM outputs, and a fun way of using @notion.com as a UI.

eointravers.com/blog/job-scr...

Data-for-good rapid response team I'm collecting people with data science skills who may want to hear about sporadic opportunities to use their skills for good on short projects. Project themes will center on topics I am connected wit...

I'm organizing a "Data for Good Rapid Response Team", i.e. a set of people who can be mobilized to work on short data science projects to help various groups and organizations. Sign up here if this is of interest to you, and please share with others with data science skills.

For one-in-three, it’s ~70.4%. One-in-four, ~68.4%. As n increases, the answer gets closer and closer to
1−1/e ≈ 63.2%, where e is Euler’s number.
en.wikipedia.org/wiki/E_(math...

Why 1-1/e​? Honestly, you would have to ask someone better at maths than me, but I think it’s a pretty cool result.

So the prob. that it does happen at least once is the probability that it *doesn’t not happen*,
1 - (1 - 1/n)^n

For a one-in-two chance, this works out as
1 - (1 - 1/2)^2 = 1 - 1/4 = 75%

The prob. of trying twice and it not happening is the prob. of it not happening the first time, times the prob. of it not happening the second time:
(1 - 1/n) * (1 - 1/n), or (1 - 1/n)^2

The prob. of it not happening in n attempts is
(1 - 1/n)^n

If you take a one-in-n chance, the probability of it coming off is 1/n​. If you roll a six-side die, the probability of rolling a 6 is 1/6​.

The prob. of the event not occurring is one minus the probability that it does occur: 1 - 1/n

But first, @xkcd.com
xkcd.com/882/

Stats time.

If you take a one-in-n chance n times (e.g. taking 10 one-in-10 chances), what’s the probability that at least one of them will come off?

Somewhat satisfyingly, the answer, regardless of what n is, turns out to be “around 63%”. Here’s why.
(Also at eointravers.com/blog/one-in-n/)

Poker is probably an interesting case study here, because AFAIK expert poker players don't try to solve K-level theory of mind problems, they just have really good heuristics.

In principle, that might mean we can get LLMs to reason under uncertainty pretty well if we fine-tune on the right heuristics?

There's an old idea in psychology (e.g. core.ac.uk/download/pdf...) that when people do perform well under uncertainty, it's because they're pattern matching using the right heuristics, rather than doing Bayesian inference.

To be fair, humans are famously also pretty bad at it, so this one might be a draw.

Humble-brag:
Over the past two years at Unmind, I got to build two AI features: Nova AI wellbeing coach, and AI practitioner matching. It's lovely to see both get a call out in Fast Company's list of the most innovative companies in the workplace for 2025.
www.fastcompany.com/91270254/wor...

Comic strip from the linked page.

Yes, but also: xkcd.com/927/

*Please repost* @sjgreenwood.bsky.social and I just launched a new personalized feed (*please pin*) that we hope will become a "must use" for #academicsky. The feed shows posts about papers filtered by *your* follower network. It's become my default Bluesky experience bsky.app/profile/pape...

a statue of a hippopotamus with its mouth open and teeth showing . Alt: A hippo being tossed a watermelon, which it crushes in its massive jaws.

I have been doing entirely too much earnest posting about deep things recently, I need to do a proper thread about hippo testicles or something just to keep myself sane.

Oh by the way hippos have migratory testicles.

and, as an ilustration, uses an LLM to quickly check the job spec against my requirements. A browser extension is a more mature way of achieving the same thing, but would take me considerably more time to get up and running.

Automation frameworks for testing like Selenium (selenium-python.readthedocs.io/index.html) are underrated as a way of quickly prototyping new AI-driven UXs. For example, this script, (<1hr work with Cursor) will open a dedicated browswer that saves every job ad I open to a database [...]

R, Snowflake, and MacOS: A match made in Hell. 😠