Igor Shilov (@_igorshilov) 's Twitter Profile
Igor Shilov

@_igorshilov

Anthropic AI Safety Fellow

PhD student at @imperialcollege.
ML, interpretability, privacy, and stuff
🏳️‍🌈

ID: 274134439

linkhttps://igorshilov.com/ calendar_today29-03-2011 20:03:53

6,6K Tweet

1,1K Followers

355 Following

Gate.io (@gate_io) 's Twitter Profile Photo
Gate.io
@gate_io

5 hours ago

🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358
Igor Shilov (@_igorshilov) 's Twitter Profile Photo
Igor Shilov
@_igorshilov

7 months ago

Hinge needs a "borderline reject" button. And a rebuttal phase

Igor Shilov (@_igorshilov) 's Twitter Profile Photo
Igor Shilov
@_igorshilov

7 months ago

Petition for the guy to change his name to Waluigi Mangione

Igor Shilov (@_igorshilov) 's Twitter Profile Photo
Igor Shilov
@_igorshilov

4 months ago

very exciting bio update

very exciting bio update
Igor Shilov (@_igorshilov) 's Twitter Profile Photo
Igor Shilov
@_igorshilov

4 months ago

SaTML in Copenhagen, here we come!

Matthieu Meeus (@matthieu_meeus) 's Twitter Profile Photo
Matthieu Meeus
@matthieu_meeus

3 months ago

And we were lucky enough to win the SaTML Conference best paper award!!🌟 Igor Shilov Yves-A. de Montjoye Marek Rei Shubham Jain Manuel Faysse

And we were lucky enough to win the <a href="/satml_conf/">SaTML Conference</a> best paper award!!🌟 <a href="/_igorshilov/">Igor Shilov</a> <a href="/yvesalexandre/">Yves-A. de Montjoye</a> <a href="/MarekRei/">Marek Rei</a> <a href="/shubhamjain0594/">Shubham Jain</a> <a href="/ManuelFaysse/">Manuel Faysse</a>
Igor Shilov (@_igorshilov) 's Twitter Profile Photo
Igor Shilov
@_igorshilov

3 months ago

Trying out o3 on Bellingcat's OSINT challenges is so fun. It's clearly close to being good at these so it's worth a shot, but not quite there yet to do it reliably. challenge.bellingcat.com It nailed an aerial photo question in just 37 seconds, but struggling with most others

AK (@_akhaliq) 's Twitter Profile Photo
AK
@_akhaliq

2 months ago

Google presents Strong Membership Inference Attacks on Massive Datasets and (Moderately) Large Language Models

Google presents Strong Membership Inference Attacks on Massive Datasets and (Moderately) Large Language Models
Ilia Shumailov🦔 (@iliaishacked) 's Twitter Profile Photo
Ilia Shumailov🦔
@iliaishacked

a month ago

Are modern large language models (LLMs) vulnerable to privacy attacks that can determine if given data was used for training? Models and dataset are quite large, what should we even expect? Our new paper looks into this exact question. 🧵 (1/10)

Are modern large language models (LLMs) vulnerable to privacy attacks that can determine if given data was used for training? Models and dataset are quite large, what should we even expect? Our new paper looks into this exact question. 🧵 (1/10)