Expert attention map

The Who's Who of AI

What credible people across AI noticed, why it matters, and where the field is converging or disagreeing.

2,397 searchable experts 4,280 tracked across all sources
Build my stream Create my edition
Filter the conversation Who is saying what?

Combine a professional role with a reaction lens. Both must match the same attributed contribution.

Clear all
New Network reaction maps

See how experts are reacting—not just what they shared.

Posts are grouped by conversation and tone. Select a lens to filter the stream; these are never permanent labels on people.

Research & technical analysis
1 live reaction map in this view Jump to the evidence ↓
Shared emphasis Neural Networks Are Already Clean Research & technical analysis · 2 attributable reactions
Showing signals surfaced by Maria Antoniak ×
Debated

The questions experts are actively pulling apart

One card per development. Sources are clustered; reaction bundles describe these posts, never the people behind them.

Established AI research Analysis 5d ago
⚡ 3 h early
neural networks already clean essay

Neural Networks Are Already Clean

11 experts across 6 network communities independently surfaced this.

11 experts 6 communities 1 sources clustered
Maria Antoniak Ted Underwood Will Oremus Ramon Astudillo Suresh Venkatasubramanian Tim Kellogg Shahab Bakhtiari Ryan Moulton +3
How the network is reacting 2 experts are emphasizing research & technical analysis.
Shared emphasis

Research & technical analysis

2 experts

Evidence, methods and technical implications.

“Quite a claim in this new work from Goodfire. Among many points, they argue that some of the main examples of superposition and polysemanticity in large models were misunderstood. This definitely deserves some deep, rigorous peer review. already-clean.goodf…”

4 experts discussed this · 14 posts
Ryan Moulton: Haven't read this yet, but I'm posting it because it seems important and I want to. already-clean.goodfire.com
Ryan Moulton: Claim, as they say.
Ryan Moulton: @norvid-studies.bsky.social some sparkling fresh juicy claims for you.
Open the full discussion →
Primary source already-clean.goodfire.com ↗
Established AI business Release 19d ago
Trillium Labs AI company launch

Introducing Trillium Labs

5 experts are actively discussing the implications.

2 experts 2 communities 1 sources clustered
Maria Antoniak Nathan Lambert

“Today we're unveiling Trillium Labs, a new non-profit to foster the open science of frontier AI. We're building open post-training recipes and will expand into open infra to study RSI, reward-hacking, multi-agent systems, and whatever comes next. blog.trill…”

5 experts discussed this · 8 posts
Nathan Lambert: Today we're unveiling Trillium Labs, a new non-profit to foster the open science of frontier AI. We're building open post-training recipes and will expand into open infra to study RSI, reward-hacki…
Nathan Lambert: At Trillium Labs, the recipes will be the slow nutrients for the seasons and the model releases will be the blooms. Building an institution dedicated to this is needed because, much as nature’s tri…
Nathan Lambert: Trilliums are wildflowers that bloom briefly in the spring, before the forest canopies fill out. Though they are small, they lay the foundation for the cycles of growth and nourishment through the …
Open the full discussion →
Primary source blog.trilliumlabs.org ↗
Developing AI research Resource 2d ago
Interfolio academic faculty job posting

Apply - Interfolio

2 experts are actively discussing the implications.

2 experts 1 community 1 sources clustered
Maria Antoniak Meredith Martin

“apply.interfolio.com/195460 new post-doc at Princeton: AI as a "cultural technology" — its role across arts & humanities & knowledge infrastructures, AI's societal effects on creative & cultural industries (literature, music, film, visual arts). Looking hum…”

2 experts discussed this · 7 posts
Meredith Martin: apply.interfolio.com/195460 new post-doc at Princeton: AI as a "cultural technology" — its role across arts & humanities & knowledge infrastructures, AI's societal effects on creative & cultural in…
Maria Antoniak: apply.interfolio.com/195460 new post-doc at Princeton: AI as a "cultural technology" — its role across arts & humanities & knowledge infrastructures, AI's societal effects on creative & cultural in…
David Smith: This looks great! Do you have any more advice for foreign graduate students? “These positions are not eligible for sponsorship of an H-1B visa requiring consular processing; other visa sponsorships…
Open the full discussion →
Primary source apply.interfolio.com ↗
Established Safety & security Research 5d ago
⚡ 4 h early
mechanistic topic models SAE paper

Model Directions, Not Words: Mechanistic Topic Models Using Sparse Autoencoders

2 experts are actively discussing the implications.

2 experts 2 communities 1 sources clustered
Maria Antoniak Jonathan Cheng
2 experts discussed this · 4 posts
Maria Antoniak: aclanthology.org/2026.tacl-1....
Maria Antoniak: babe wake up, new results just dropped, lda is still great
Ted Underwood: babe wake up, new results just dropped, lda is still great
Open the full discussion →
Primary source aclanthology.org ↗
More conversations

What experts are discussing without an anchoring article

9 experts · 12 posts · 17d ago
Maria Antoniak Eugene Vinitsky David Pfau joao Thomas Steinke Shahab Bakhtiari David Picard Marc Deisenroth
David Pfau: An academic collaborator just told me his students aren't using LLMs because he can't afford even the low tier monthly subscription for everyone. I'm not sure the folks making $Xm/yr at frontier la…
ae: An academic collaborator just told me his students aren't using LLMs because he can't afford even the low tier monthly subscription for everyone. I'm not sure the folks making $Xm/yr at frontier la…
Open the thread →
4 experts · 8 posts · 13d ago
Maria Antoniak David Smith Henry Farrell Hoyt Long
Maria Antoniak: This talk by @himself.bsky.social was wonderful. Models mechanize, map, + remix the "house of intellect", connecting together previously disconnected hallways of knowledge. LLMs are not stochastic …
David Smith: This talk by @himself.bsky.social was wonderful. Models mechanize, map, + remix the "house of intellect", connecting together previously disconnected hallways of knowledge. LLMs are not stochastic …
Open the thread →
4 experts · 5 posts · 18d ago
Maria Antoniak Ted Underwood Christian Wolf Scott McGrath
Multiple readings Concern & critique · 2 Questions & unknowns · 1
Maria Antoniak: Maybe worth saying aloud that I do worry about safety risks (hacking, bio, an “out of control” agent). I just also see those risks as deeply entangled with money, politics, personalities, regulatio…
Ted Underwood: Maybe worth saying aloud that I do worry about safety risks (hacking, bio, an “out of control” agent). I just also see those risks as deeply entangled with money, politics, personalities, regulatio…
Open the thread →
3 experts · 5 posts · 9d ago
Maria Antoniak Ted Underwood Dallas Card
Multiple readings Research & technical analysis · 1 Questions & unknowns · 1
Maria Antoniak: I feel like there is a pattern in #CHR2027 papers of "we engineered this LLM-based pipeline and evaluated it in service of a humanities research question, but for CHR we will only report about the …
Ted Underwood: I feel like there is a pattern in #CHR2027 papers of "we engineered this LLM-based pipeline and evaluated it in service of a humanities research question, but for CHR we will only report about the …
Open the thread →
2 experts · 6 posts · 4d ago
Maria Antoniak Ted Underwood
Ted Underwood: this is very exciting paper, but for me the biggest insight is that classic LDA still goes hard as a topic model. using SAE features as inputs to topic models is uncanny, like strapping a vivisecte…
Ted Underwood: what's kind of fascinating today is that "interpreting topic models," "interpreting neural networks," and "topic modeling neural networks" are all suddenly in my feed, and you + Nick Cammarata are …
Open the thread →
3 experts · 4 posts · 19d ago
Maria Antoniak Maxim Raginsky David Picard
Maxim Raginsky: ... by the vigilant reader, who is not unaware of the fact that she brings the text to meaningfulness thanks to its lacunae, whether these be intended or not." (footnote 23 on p. 159, _Oneself As A…
Maxim Raginsky: Ricoeur on reading, very relevant for LLMs that act via the interface of language: "reading ... is at its best a struggle between two strategies, the _strategy of seduction_ pursued by the author i…
Open the thread →
2 experts · 5 posts · 8d ago
Maria Antoniak angela zhou
Multiple readings Building & implementation · 1 Opportunity & adoption · 1
angela zhou: If I weren't underwater on X revisions I'd give it a go, but happy to link up with folks. There's probably a real value add there. To quote an old friend who was catching up with AI terms and "coul…
angela zhou: if any of y'all have any friends at these places ... i pledge to them at least my bluesky engagement for life. Maybe the capacity to build on atproto can support a feed / a higher-signal discussion…
Open the thread →