A. Feder Cooper
@afedercooper.bsky.social
📤 373
📥 202
📝 159
ML researcher, Stanford postdoc affiliate, future Yale professor
https://afedercooper.info
reposted by
A. Feder Cooper
Kari Maaren
11 days ago
The whole point of being an academic is that you need to be willing to spend three days creating a 700-word footnote that you will later delete. And you need to LIKE IT.
24
904
192
[NeurIPS '25] Our poster (1110) for “Comparison requires valid measurement: Rethinking attack success rate comparisons in AI red teaming ” is on Friday, December 5, 4:30pm-7:30pm PST in Exhibit Hall C,D,E. [https://openreview.net/forum?id=d7hqAhLvWG]
26 days ago
1
2
0
I’ll be hanging out at our poster on membership inference, but in the same slot Brian Lester will present our work on “The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text” (poster 102)! [https://arxiv.org/abs/2506.05209]
add a skeleton here at some point
26 days ago
1
4
2
[NeurIPS '25] Really excited to present “Exploring the limits of strong membership inference attacks on large language models” (poster 1300) this morning (Friday December 5, 11am-2pm in Exhibit Hall C-E)! [https://arxiv.org/abs/2505.18773]
26 days ago
1
2
1
[NeurIPS '25] Our oral slot and poster session on "Machine Unlearning Doesn't Do What You Think: Lessons for Generative AI Policy and Research" are tomorrow, December 4! [https://arxiv.org/abs/2412.06966] Oral: 3:30-4pm PST, Upper Level Ballroom 20AB Poster 1307: 4:30:-7:30pm PST, Exhibit Hall C-E
28 days ago
1
2
2
Tutorial tomorrow at 1:30PM PST! My talk slots will cover memorization + copying in models and their outputs, canonical extraction methods, and recent work with
@marklemley.bsky.social
and others on extracting pieces of memorized books from open-weight models.
arxiv.org/abs/2505.12546
add a skeleton here at some point
30 days ago
0
9
4
reposted by
A. Feder Cooper
Katherine Lee
about 1 month ago
I'm at NeurIPS & hiring for our pretraining safety team at OpenAI! Email me if you want to chat about making safer base models!
2
4
2
Excited to be at NeurIPS this week in San Diego! Please reach out (best over email) if you’d like to chat about privacy & security, scalable evals, and reliable ML systems. I’ll be presenting a few papers/speaking at some events, please stop by! Will post details throughout the week (summary below)
about 1 month ago
1
7
1
reposted by
A. Feder Cooper
Johan Ugander
about 1 month ago
📣 Postdocs at Yale FDS! 📣 Tremendous freedom to work on data science problems with faculty across campus, multi-year, great salary. Deadline 12/15. Spread the word! Application:
academicjobsonline.org/ajo/jobs/31114
More about Yale FDS:
fds.yale.edu
loading . . .
Yale University, Institute for the Foundations of Data Science
Job #AJO31114, Postdoc in Foundations of Data Science, Institute for the Foundations of Data Science, Yale University, New Haven, Connecticut, US
https://academicjobsonline.org/ajo/jobs/31114
0
23
14
Just finished reading the GEMA v. OpenAI decision (slowly, my German isn't great). Looks like a not small part of the analysis tracked parts of arguments
@jtlg.bsky.social
and I made in 2024. I don't have a well-formed response yet, but hopefully soon. (Main thought atm is a very unpolished "woah")
add a skeleton here at some point
about 2 months ago
0
3
0
reposted by
A. Feder Cooper
James Grimmelmann
about 2 months ago
Today's decision in GEMA v. OpenAI by a German court holds that ChatGPT infringes copyright when it memorizes song lyrics. The opinion cites my paper with
@afedercooper.bsky.social
on memorization in generative models, and its analysis tracks ours.
drive.google.com/file/d/1dUaD...
loading . . .
42-O-14139-24-Endurteil.pdf
https://drive.google.com/file/d/1dUaDiRoPG5v7R7UxNQzEM31yS9pWsknm/view
1
33
18
reposted by
A. Feder Cooper
Mina Kimes
about 2 months ago
Bill Ackman gotta be on the third draft of a tweet longer than Middlemarch right now
224
12403
1240
I’m kinda known as a copyright person, but (even in memorization) I mainly study how to draw reliable conclusions from large-scale AI/ML systems. There’s a long spiel why, but today I feel defeated. 100 hours/week on this for 6 years, just to find out a parent treats Gemini in search as ground-truth
about 2 months ago
0
4
0
The NeurIPS position track didn't take a large number of extraordinary papers that surpassed the acceptance bar, limiting the acceptance rate to an unusually low 6%. If you have a rejected paper at the intersection of ML and law, consider submitting to ACM CSLaw '26.
loading . . .
2026-CFP - ACM Symposium on Computer Science & Law
2026 Call for Papers 5th ACM Symposium on Computer Science and Law March 3-5, 2026 Berkeley, California The 5th ACM…
https://computersciencelaw.org/2026-2/2026-cfp/
3 months ago
1
5
2
reposted by
A. Feder Cooper
Mark Lemley
3 months ago
Our paper "Machine Unlearning Doesn't Do What You Think" was accepted for presentation at NeurIPS Congrats
@afedercooper.bsky.social
and
@katherinelee.bsky.social
, who led the effort
arxiv.org/abs/2412.06966
loading . . .
Machine Unlearning Doesn't Do What You Think: Lessons for Generative AI Policy, Research, and Practice
We articulate fundamental mismatches between technical methods for machine unlearning in Generative AI, and documented aspirations for broader impact that these methods could have for law and policy. ...
https://arxiv.org/abs/2412.06966
1
21
4
One more week to submit to CSLaw '26!!
add a skeleton here at some point
3 months ago
0
0
0
reposted by
A. Feder Cooper
Pamela Samuelson
4 months ago
For an update on the state of play in the generative AI copyright cases, try this podcast:
shows.acast.com/arbiters-of-...
loading . . .
AI Copyright Lawsuits with Pam Samuelson | Scaling Laws
https://shows.acast.com/arbiters-of-truth/episodes/ai-copyright-lawsuits-with-pam-samuelson
0
6
1
15 days left to submit to the CSLaw '26 main track! (archival and non-archival)!
add a skeleton here at some point
4 months ago
0
5
5
reposted by
A. Feder Cooper
Elizabeth Lopatto
4 months ago
did a little media criticism
www.theverge.com/politics/777...
loading . . .
The WSJ carelessly spread anti-trans misinformation
The Wall Street Journal’s fuckup while covering Charlie Kirk’s killing needs more than an editor’s note.
https://www.theverge.com/politics/777630/wsj-trans-misinformation-charlie-kirk
29
2608
659
reposted by
A. Feder Cooper
mattie lubchansky
4 months ago
was just looking for
@seantcollins.com
’s “goofy at the crucification” post and google is so cool now
11
321
45
After 2 years in press, it's published! "Talkin' 'Bout AI Generation: Copyright and the Generative-AI Supply Chain," is out in the 72nd volume of the Journal of the Copyright Society
copyrightsociety.org/journal-entr...
written with
@katherinelee.bsky.social
&
@jtlg.bsky.social
(2023)
loading . . .
TALKIN' 'BOUT AI GENERATION: COPYRIGHT AND THE GENERATIVE-AI SUPPLY CHAIN | The Copyright Society
We know copyright
https://copyrightsociety.org/journal-entries/talkin-bout-ai-generation-copyright-and-the-generative-ai-supply-chain/
4 months ago
1
14
4
reposted by
A. Feder Cooper
James Grimmelmann
4 months ago
The Bartz v. Anthropic settlement is the polar opposite of the Google Books settlement: a discrete one-time payment for past copying, on a discrete and closed-ended class, and making no attempt at all to deal with a larger forward-looking issues.
0
19
5
reposted by
A. Feder Cooper
Mark Lemley
4 months ago
Here is the direct link to the paper:
arxiv.org/abs/2505.12546
add a skeleton here at some point
0
12
5
I’m excited to share that my paper with
@jtlg.bsky.social
, "The Files are in the Computer: On Copyright, Memorization, and Generative AI" (April 2024), is out in the AI Disrupting Law symposium issue of the Chicago-Kent Law Review! The full issue is here:
scholarship.kentlaw.iit.edu/cklawreview/
loading . . .
Chicago-Kent Law Review | Chicago-Kent College of Law
https://scholarship.kentlaw.iit.edu/cklawreview/
4 months ago
1
23
5
The CFP for ACM CSLaw '26 is up! Deadline for main-track papers (archival and non-archival) is September 30!
computersciencelaw.org/2026
loading . . .
2026 - ACM Symposium on Computer Science & Law
CS&Law 2026 5th ACM Symposium on Computer Science and Law March 3–5, 2026 Berkeley, California Computing, software, and the Internet…
https://computersciencelaw.org/2026
5 months ago
0
12
10
I understand what the underlying probabilities mean, and therefore why this was worth giving a go. But I’m still occasionally like “How tf can someone extract entire books from a frontier company’s flagship LLM? Like we got _all_ of HP 1 with just ‘Mr. and Mrs. D’ as the seed prompt? What??”
5 months ago
0
2
0
Had a great time and learned a ton at ICML. But as an introvert, I’ve used up all my talking budget until the fall. Excited to get back to full time researchy things, and will hopefully have some exciting new results to share soon!
5 months ago
1
4
0
reposted by
A. Feder Cooper
M A Osborne
6 months ago
Strangers love to tell me “I can’t understand you, because of your MASK”. Dude, I am literally someone who gets paid to speak to large audiences while wearing a mask—I know I can be understood!
14
148
15
Happening now! Please swing by to talk about measurement!
add a skeleton here at some point
6 months ago
0
2
0
Excited to be at
#ICML
'25! Please reach out if you'd like to chat. You can also find me presenting work at a few different spots, listed below!
6 months ago
2
2
0
Feeling so excited + grateful to be representing this paper at
#ICML
! Please stop by to talk about how to do more valid measurement for evaling gen AI systems! Work led by the incomparable
@hannawallach.bsky.social
and
@azjacobs.bsky.social
as a part of Microsoft’s AI and Society initiative!!
add a skeleton here at some point
6 months ago
0
12
2
Some minor updates to our recent books memorization paper! I’ve separated out a new section 5 that I hope makes some of our ML findings about memorization clearer to a wider audience. Preprint here:
arxiv.org/abs/2505.12546
1/8
loading . . .
Extracting memorized pieces of (copyrighted) books from open-weight language models
Plaintiffs and defendants in copyright lawsuits over generative AI often make sweeping, opposing claims about the extent to which large language models (LLMs) have memorized plaintiffs' protected expr...
https://arxiv.org/abs/2505.12546
6 months ago
1
2
1
reposted by
A. Feder Cooper
Riana
6 months ago
"Llama 3.1 70B memorizes some books, like Harry Potter & the Sorcerer's Stone and 1984, almost entirely. ... HP is so memorized that, using a seed prompt consisting of just the first line of chapter 1, we can deterministically generate the entire book near-verbatim."
papers.ssrn.com/sol3/papers....
loading . . .
Extracting memorized pieces of (copyrighted) books from open-weight language models
Plaintiffs and defendants in copyright lawsuits over generative AI often make sweeping, opposing claims about the extent to which large language models (LLMs) h
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=5262084
0
6
5
reposted by
A. Feder Cooper
Blake E. Reid
6 months ago
Once again, I encourage folks speculating about what this means to read
@pamelasamuelson.bsky.social
on remedies. The range of possibilities is quite broad.
loading . . .
Thinking About Possible Remedies in the Generative AI Copyright Cases
The sixteen lawsuits brought to date against OpenAI and other developers of generative AI technologies include claims that making copies of in-copyright works f
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4770671
1
28
3
reposted by
A. Feder Cooper
Blake E. Reid
6 months ago
This opinion is a reminder that these cases are not general-purpose referenda on AI policy; they are hyper-technocratic copyright cases. Copyright draws lots of unsatisfying and counterintuitive distinctions, which is why you should hire and listen to copyright lawyers on the front end.
1
40
8
reposted by
A. Feder Cooper
Luis Villa
6 months ago
“these are hypertechnocratic” is one of the most important things you can draw from this morning’s ruling. In other words, hesitate before drawing parallels between this case and your most (loved|hated) AI training use case. (
@chup.blakereid.org
’s whole thread is great)
add a skeleton here at some point
0
9
1
very proud of my hometown. stunned, but proud.
6 months ago
1
0
0
idk folks, have read today’s decision by Alsup and I think it’s narrower than news headlines are indicating (shocker).To me this reads like (mostly) a win for Anthropic, but in this specific case with respect to these claims from these specific plaintiffs.
6 months ago
0
2
1
reposted by
A. Feder Cooper
Michael Hobbes
6 months ago
The conservative movement is lying about basically every empirical "debate" of our time, from climate change to voter fraud to vaccines. Why is it so impossible for reactionary centrists to contemplate the possibility they're lying about trans healthcare too?
add a skeleton here at some point
82
4835
1291
reposted by
A. Feder Cooper
Mark Lemley
6 months ago
Ars Technica has a great writeup of my new study with
@afedercooper.bsky.social
and Amy Cyphert on memorization in AI language models
arstechnica.com/ai/2025/06/s...
loading . . .
Study: Meta AI model can reproduce almost half of Harry Potter book
The research could have big implications for generative AI copyright lawsuits.
https://arstechnica.com/ai/2025/06/study-metas-llama-3-1-can-recall-42-percent-of-the-first-harry-potter-book/
0
13
1
reposted by
A. Feder Cooper
Dr. Casey Fiesler
6 months ago
I was wondering when tech media was going to pick up this paper. :) "Study: Meta AI model can reproduce almost half of Harry Potter book":
arstechnica.com/features/202...
Coincidentally, I made a video about this paper a couple of weeks ago in a response to a comment I got elsewhere online.
loading . . .
1
35
8
reposted by
A. Feder Cooper
Ars Technica
6 months ago
For AI industry critics, the big takeaway here is that—at least for some models and some books—memorization is not a fringe phenomenon.
loading . . .
Study: Meta AI model can reproduce almost half of Harry Potter book
The research could have big implications for generative AI copyright lawsuits.
https://arstechnica.com/features/2025/06/study-metas-llama-3-1-can-recall-42-percent-of-the-first-harry-potter-book/?utm_source=bluesky&utm_medium=social&utm_campaign=aud-dev&utm_social-type=owned
9
41
20
uh...lol(?) at all the internet commenters who think that we/the authors on our recent books extraction paper are tacitly endorsing or supporting Rowling because Harry Potter is one of the books we tested (that happens to be highly memorized)
6 months ago
0
5
0
reposted by
A. Feder Cooper
Jessica Clarke
7 months ago
Today's Skrmetti opinion is devastating to transgender children and families who live in states with cruel laws barring gender affirming care. But it is very important to recognize this opinion does not give private entities, legislatures, or the President carte blanche to discriminate! 1/x
5
431
166
excited to see that folks are finding their way to our books memorization project.
mashable.com/article/meta...
I’ll also be giving a talk on this work at the ICML memorization workshop in July. Please check out the paper for more details!
arxiv.org/abs/2505.12546
cc
@marklemley.bsky.social
loading . . .
Meta's AI tool Llama 'almost entirely' memorized Harry Potter book, study finds
Llama has a memorization problem, but don't call it a smoking gun.
https://mashable.com/article/meta-llama-reproduce-excerpts-harry-potter-book-research
7 months ago
0
10
3
reposted by
A. Feder Cooper
Techmeme
7 months ago
Researchers find Llama 3.1 recalls large parts of popular copyrighted books, possibly weakening AI industry claims that such memorization is fringe behavior (Timothy B. Lee/Understanding AI)
Main Link
|
Techmeme Permalink
4
53
25
reposted by
A. Feder Cooper
Chad Loder
7 months ago
Gavin Newsom vetoed a data broker privacy bill that our legislature passed last year that would have allowed Californians to opt out of all data brokers with a single click.
add a skeleton here at some point
31
5210
1469
Tim Lee wrote an awesome short piece on my recent project with
@marklemley.bsky.social
and others on extracting memorized books from LLMs. Please check it out for a very clear explanation of what we did, and why the results have important consequences!
www.understandingai.org/p/metas-llam...
loading . . .
Meta's Llama 3.1 can recall 42 percent of the first Harry Potter book
New research could have big implications for copyright lawsuits against generative AI.
https://www.understandingai.org/p/metas-llama-31-can-recall-42-percent
7 months ago
0
11
3
reposted by
A. Feder Cooper
Rebecca Tushnet
7 months ago
Licensing fee demand incoming
add a skeleton here at some point
0
15
6
wow, a research career milestone: someone angrily saying my paper has been done before…by pointing to my prior work that does something different (and which this current work builds on / depends on)
7 months ago
0
1
0
Load more
feeds!
log in