I am delighted to share that I'll be joining the University of Washington Information School as an Assistant Professor!

@kennypeng.bsky.social ultimate traversle.io x skytrails.org mashup

Dan Hon@danhon.com · 7d ago

something something bluesky odyssey game drops you at a rando post and you have to get back to your own timeline somehow

It was great to have the chance to share about this work led by @mariannealq.bsky.social on AI overviews in local news environments… keep an eye out for more on this soon! #ic2s2

Manoel Horta Ribeiro@manoelhortaribeiro.bsky.social · last wk.

@mariannealq.bsky.social and @isabelcorpus.bsky.social presenting their work on local variations on AI overviews! They find that AI overviews are much better in high information environments.

At my first #IC2S2 this week -- would be great to meet and chat with folks here!

Here's a feed for IC2S2! It's a reverse chronological feed that matches all string mentions of "IC2S2", the hashtags #IC2S2 and #IC2S22026, plus all mentions of and posts by @ic2s2.bsky.social. 📌 Pin the feed to follow along with the conference this week!

Come join us in Toronto. UofT is hiring an Associate Professor/Professor and endowed Hinton Chair in Artificial Intelligence. @utoronto.ca @uoftcompsci.bsky.social jobs.utoronto.ca/job/Toronto-...

I'm at @ic2s2.bsky.social this week! I'm looking forward to presenting Paper Skygest on Thursday afternoon (Social Media Platforms and Online Communities) If you're around, I'd love to chat about social media research on Bluesky / beyond!

Looking forward to giving a tutorial on 𝗦𝗼𝗰𝗶𝗮𝗹 𝗠𝗲𝗱𝗶𝗮 𝗙𝗲𝗲𝗱 𝗥𝗮𝗻𝗸𝗶𝗻𝗴 𝗔𝗹𝗴𝗼𝗿𝗶𝘁𝗵𝗺𝘀 with @tiziano.bsky.social at @ic2s2.bsky.social! 🕜 𝗧𝗶𝗺𝗲: Tuesday, 1-4 pm 📍 𝗟𝗼𝗰𝗮𝘁𝗶𝗼𝗻: Mansfield (210). 🌐 𝗪𝗲𝗯𝘀𝗶𝘁𝗲: social-media-ais.github.io/ic2s2-26-tut...

fwiw distillation was also illegal back in 1926 too, so i guess history repeats

We have re-opened our search for a 3-year AI-research postdoc at NYU's Center for Social Media, AI, and Politics!! The position starts this **September 1st** (or as soon as possible thereafter) and is in person at NYU. If necessary, we will also consider a 1/1/27 start. 1/3

Wow 跟我一样

John Lake@jlake9.bsky.social · 2w ago

1/ Neel Nanda reports a wild interp finding: when a model is confused about a sentence’s meaning, Chinese “what does this mean?” characters can surface in latent J-Space. https://x.com/NeelNanda5/status/2079676406053961754

People never learn. Unless they're on LinkedIn, then they learn all the time and the lessons are everywhere. People never change. Unless they're on LinkedIn, then everything does so "completely."

Principle component analysis

I’m joining Cornell as an Assistant Professor in Computer Science this fall! My group will work on probabilistic machine learning for scientific discovery. I’m recruiting PhD students to start in Fall 2027 -- and stay tuned for postdoc opportunities!

I will literally follow any and all ML PhD students on this website, where you at

tomorrow (today over in Korea) Haruka will be presenting this paper in the ICML. working on this project with her (and folks at Meta) is part of what convinced me to revamp my undergraduate RL course (more on that later). check out the poster at the Wednesday poster session! arxiv.org/abs/2605.26385

Credit-assigned Policy Gradient for Early Stage Retrieval in Two-stage Ranking

Large-scale search, recommendation, and retrieval-augmented generation (RAG) systems typically employ a two-stage architecture: an early-stage ranker (ESR) generates a candidate set, which is subseque...

arxiv.org

Are you at ICML next week? Feel like your decision-making for which sessions to attend might not be risk minimizing? Don't incur (swap) regret and come to my, @aaroth.bsky.social, and @ncollina.bsky.social's tutorial Monday on multicalibration, decision-making, and collaborative learning!

1/7 Excited to share EconCSLib: a Lean library/workflow. Our vision is to enable researchers who don't know Lean to formalize their Applied Modeling papers. Why? I, and most theorists, are terrified of fatal mistakes in our papers Paper: arxiv.org/abs/2606.13306 Project: gargnikhil.com/EconCSLib/

why have all platforms standardized on a binary like/not like signal with very rare "dislike" or "superlike"

Sophie Greenwood@sjgreenwood.bsky.social · last mo.

yeah i've been thinking about this question lately, is it e.g. because it's not a helpful signal for maximizing the metrics they care about or because there aren't many ways of doing this without adding friction etc...

what if... top chicken prediction market as a content recommendation engine: incentivized community curation of engaging content? (interesting dynamics here: even assuming good faith behavior, predictions change the rankings, which changes exposure/engagement, etc)

Singularity's Bounty e/CC@catblanketflower.yuwakisa.com · last mo.

I actually checked the top chicken prediction market before I checked Bluesky today 🫤 topchicken.cee.wtf#/

Presenting this at #COLT2026! Let me know if you want to chat about (1) the linear representation hypothesis, (2) theoretical foundations for interpretability / science of AI. - Talk: Thursday 2:33 pm - Deep Learning Theory / Optimization - Poster: Thursday Lunch)

Kenny Peng@kennypeng.bsky.social · 6mo ago

New paper! The Linear Representation Hypothesis is a powerful intuition for how language models work, but lacks formalization. We give a mathematical framework in which we can ask and answer a basic question: how many features can be stored under the hypothesis? 🧵 arxiv.org/abs/2602.11246

Are you a Bluesky user or developer? Princeton's HCI group is conducting an IRB-approved interview study on perspectives about generative AI (genAI) in decentralized social media. Participation includes: • 45–60 min Zoom interview • $30 USD digital gift card Signup: forms.gle/EDmcLnjbamtb...

1/ Privacy-preserving credentials can prove you're over 18, but not much else. Our new research changes that. 𝜋Creds introduces verifiable credentials generated by trusted LLM inference over authenticated.

Our paper with formal analysis and code: arxiv.org/pdf/2606.03771 H/T to my co-authors Sam Breckenridge, Derek Leung, Andrés Fábrega, James Austgen, Farinaz Koushanfar, and Prof. Ari Juels

I wrote a custom Bluesky feed! It's a classifier trained my past likes. As a bonus, I got to learn more about recommender systems ("RecSys"). adamwiggins.com/posts/a-blu...

A Bluesky feed for one · Adam Wiggins

I built a custom feed for Bluesky that trains a model on my past interactions. Here's what I learned about RecSys and social media feeds generally.

adamwiggins.com

The list of 290 papers we accepted for EC'26 is now up: ec26.sigecom.org/program/acce... We are really grateful to the whole programme committee for their hard work and to the authors for submitting many terrific papers. See you all in Rome in July 🍕🇮🇹🍝☀️🤖

spotted an ad for @devingaffney.com on the nyc subway

Giving a "pub talk" tomorrow evening thanks to @pintofscience.us. (And folks talking all week, e.g. see @informor.bsky.social on Wednesday). www.pintofscience.us/event/who-de...

Who Decides? Law, Values, and Algorithms in an Automated World

From courtrooms to code, how much should we trust algorithms? Explore why judges using LLMs may be misguided, how values are socially constructed, and how …

pintofscience.us

Mor Naaman@informor.bsky.social · 3mo ago

Hey NYC friends, this is something really cool and unusual that I am doing in a couple of weeks -- a public pub talk about my research! Thanks to @pintofscience.us for organizing. Wednesday, May 20th at 7pm (it's $2 plus the beer you will be having). pintofscience.us/event/smart-...

We are very excited to announce our first workshop on From Theory to Practice: behind the scenes on research deployments at EC’26 (July 6 in Rome)! Call for posters and submissions now open! Organized by myself, @ericachiang.bsky.social , Bailey Flanigan, @brwilder.bsky.social

I got from SATURDAY to DIRTY in 4 steps in today's Traversle puzzle! Play here: traversle.io