Just wanna remind everyone that this is a pro science account and I will not be tolerating science denialism

McCarthyism is real. I’ve been relieved of teaching responsibilities. Don’t stop talking about Palestine.

Being trans means that you can no longer have privacy and your private medical records are subject to the whims of politicians who don't want you to exist. Tennessee, Texas, Missouri, and the UK have all demanded access of trans patient records for political purposes.

OK, one of the (many) things that irks me with that NeurIPS 'High School projects' track:
You shouldn't set the expectation for HS students to have papers, or research experience.
Hell, you shouldn't set the expectation for UNDERGRADS to have papers, or research experience.

There's a pattern that predicts who does well on IQ tests: people who spend more time on hard problems.

New work w/ Sam Cheyette extends a 40 year old finding about the role of effort in IQ testing, showing why these tests aren't bare measures of 'intelligence.'

CLeaR conference was an absolute blast!

45 posters, 12 oral presentations, 5 keynotes...

1/n

#causality #causalAI #causalinference #machinelearning #conference #causaltwitter CLeaR-Conference on Causal Learning and Reasoning Judea Pearl

for a large sample size, it's better to interpret confidence intervals than use p-values ('significant' effects may be practically meaningless)

for a small sample size, it's better to interpret confidence intervals than use p-values (emphasises the uncertainty in results)

Metaphysics is gross

Truly outstanding paper on how people create new concepts by building on simpler ones.

I used to think this was an esoteric philosophical topic, but this new computational account, and the experimental effects it predicts, are just super compelling.

🚨Pre-print alert:🚨
Have we built machines that think like people?
In new work, led by Luca Schulze Buschoff and Elif Akata and together with Matthias Bethge, we assess multi-modal #LLMs reasoning abilities in three core domains: intuitive physics, causality, and intuitive psychology.

How do people come to opposite causal judgments about whether a public health policy worked?

We find that, even if people agree on what actually happened (i.e., viruses decreased), they can be made to disagree about the *counterfactual* of what would have happened otherwise.

'AI is just linear algebra' feels like a midwit thing, where someone's figured out enough about the world to be excited by reductionism, but not enough to realise that it's universal.

The state of benchmarks in AI is really bad.

In case you are wondering, Karthik Valmeekam found that LLM's still can't plan even after the advent of Claude 3 Opus and Gemini Pro..😬

tldr; no AGI doom by all fools day! You can still tame the lot by throwing assorted blocks at 'em..😋

The much ballyhooed Claude 3 Opus does no…

Circuits are a hot topic in interpretability, but how do you find a circuit and guarantee it reflects how your model works?

We (Sandro Pezzelle, Yonatan Belinkov, and I) introduce a new circuit-finding method, EAP-IG, and show it finds more faithful circuits arxiv.org/abs/2403.17806 1/8

This is one of my favorite posts on CrossValidated because it clearly describes what each assumption in linear regression means and what it is necessary and sufficient for. A great antidote to people thinking residuals have to be normally distributed.

stats.stackexchange.com/a/16460/116195

Do *additional* pubs in lower-ranked journals help or hurt a CV?

1. In psych, help: psycnet.apa.org/record/2019-27…

2. In econ, hurt*: CV with A top pubs rated better than A top + B non-top😆
sciencedirect.com/science/articl…

* small sample, depends on if joint-eval or not, etc.

Super cool not only because of the catchy findings, but also how this points out a way of examining LLMs using the *methodology* of cognitive science

This paper is super cool! My favorite new finding in causal generalization (if I may 🥰)

Jules Matteo Capucci I was looking to characterise algebraic topology as a research programme, using something rather slogan-like rather than the axioms that Lakatos had chosen.