Ian Stevenson

@UConnPSYC // Neuroscience and Nerdery // @ihstevenson on other social media too

[Ian Stevenson @ihstevenson]

New preprint with work from @AlexCarmenClo modeling speech recognition in natural noise 👂💬🤖🧠🤯 https://t.co/lxAE3nKU7D

[Jieyu Zheng @jieyusz]

Nice Ambition from @elonmusk @djseo_ @neuralink 🚀, but we @mameister4 kindly invite you to read our words of caution first😉⚠️, now on arXIV: https://t.co/rUUyNaWoD1

Academics, please join the discussion on alphaxiv: https://t.co/IHaA4FhPvg. Just one thought at a time please!

[Merriam-Webster @MerriamWebster]

TLC | abbreviation | tender loving care

[Jonathan Pillow @jpillowtime]

In 2019 I went to Cambridge to debate @lengyel_m on the usefulness of Marr's Three Levels of Analysis. We've finally published a pair of "CrossTalk" articles in @JPhysiol laying out our positions.

My piece: https://t.co/0RWNM9Hk54

#MarrNoMore #DownWithThe3Levels

[Ian Stevenson @ihstevenson]

This model also gives an interpretable "transfer function" for participant behavior, and, interestingly, we found that humans and neural network models are sensitive to different features 4/

[Ian Stevenson @ihstevenson]

Alex found that some background sound features improve word recognition while others hurt it. Recognizing foreground speech in background speech babble is much harder than listening in white noise, even at the same signal-to-noise ratio, and even after the spectra are matched 1/

[Ian Stevenson @ihstevenson]

New preprint with work from @AlexCarmenClo modeling speech recognition in natural noise 👂💬🤖🧠🤯 https://t.co/lxAE3nKU7D

[Ian Stevenson @ihstevenson]

Alex's model goes beyond "energetic masking" and shows that the modulation statistics play a major role, and a low-dimensional model explains ~90% of the variance in behavior 3/

[Brad Simpson (bradleyrsimpson.bsky.social) @bradleyrsimpson]

Our state historian, Andy Horowitz: Divesting from CT's colleges abandons hope for the future. https://t.co/5WO57yaYRa

[Alex Clonan @AlexCarmenClo]

Check out our new preprint! We investigate how spectrum & modulation in natural sounds can help/hinder human speech recognition👂💬

Further- we explore the bottom-up acoustic invariances driving this effect with a midbrain inspired, interpretable model🧠

https://t.co/SyWrKenCg0

[Latif Nasser @latifnasser]

Last January, I noticed something peculiar in my 2yo’s bedroom that - after a year of obsessive reporting - led me to a profound cosmic revelation about what’s even possible in our universe. A 🧵.

[Richard Gao @_rdgao]

I'm presenting at NeurIPS next week, lots of things to say about this paper, but 2 things I'm most proud of is that multiple reviewers said, despite the very complicated title:

1. it's well-written, and
2. "the idea is simple and straightforward, but not in a bad way" (!?)

[Patrick Mineault @patrickmineault]

I almost DDOS'ed myself by asking you for your NeurIPS papers! I found 197 NeuroAI and NeuroAI-adjacent papers at the main conference–brace yourselves! https://t.co/gVBNAU8REC

[Playfool @studioplayfool]

NEW PROJECT: How (not) to get hit by a self-driving car 🚗⚠️💥

A street-based game that challenges visitors to avoid being detected by an AI-powered camera. Can you fool the AI and reach the goal?

Collab with @tomokihara as part of @wshed @PlayableCity

[Ian Stevenson @ihstevenson]

@CellTypist Yes! Thanks for sharing that great dataset

[Ian Stevenson @ihstevenson]

We have a new preprint out with some progress on the puzzle of why many Bayesian decoders are overconfident... https://t.co/XzL0TvWchI

[Ian Stevenson @ihstevenson]

and M1...

[Ian Stevenson @ihstevenson]

including some results relating posterior uncertainty to decoder accuracy in V1...

[Ian Stevenson @ihstevenson]

We have a new preprint out with some progress on the puzzle of why many Bayesian decoders are overconfident... https://t.co/XzL0TvWchI