Lucile Ter-Minassian

Senior Research Scientist @Owkin, Dr in Stats & ML @UniofOxford. Former @IBMResearch and @GoogleAI intern

[Lucile Ter-Minassian @TerLucile]

great work!

[Seraphina Goldfarb-Tarrant @seraphinagt]

New paper from @Preethi__S_ internship @cohere! https://t.co/7jNYXBuOQv
LLMs are very commonly used in HR, but most fairness work on it in a generative context is contrived and unrealistic. We build fairness tests based on observed real usage with super interesting results 🧵

[Daniel Paleka @dpaleka]

What happened recently in AI/ML safety research (1/8) 🧵:

[Lucile Ter-Minassian @TerLucile]

is anyone else also constantly scrolling on their endless Claude/ChatGPT thread? yo @OpenAI @AnthropicAI a file outline we can click on to get back to previous questions would be great

[Lucile Ter-Minassian @TerLucile]

the paper is out! https://t.co/TUNeiNEtPX

[Lucile Ter-Minassian @TerLucile]

Do LLMs plan the content of a paragraph at its onset? Together with Nicholas Pochinkov, Angelo Benoit,
@lovkushatleeds, and Zainab Ali Majid, we take a mechanistic interpretability approach to investigate this research question.
https://t.co/V8GL4lFxoL

[Lucile Ter-Minassian @TerLucile]

Our findings:
By transferring the activations from the final layer (amongst others), we are hinting at what the next token is. We thus added 'cheat' tokens to the neutral baseline.

(2) neutral + 2 tokens compares with our transferred generations wrt closeness to original

[Lucile Ter-Minassian @TerLucile]

🥹🥹🥹🥹🥹🥹

[Lovkush Agarwal 🔸 @lovkushatleeds]

My first paper in AI Safety now on arxiv. https://t.co/OU4Qahw15m

Big thanks to my teammates and to the SPAR orga using team. Huge credit goes to Lucile for doing the heavy lifting with writing and logistics for the paper.

[Lucile Ter-Minassian @TerLucile]

started @BlueDotImpact's AI Governance course today! Really impressed with the quality and organisation. Excited to learn about policies that improve AI safety

[Lucile Ter-Minassian @TerLucile]

(1/2) as a side hustle: I've decided to start a series, writing reviews on Mechanistic Interpretability papers. I'm mainly going to sample from @NeelNanda5's v2 list

mech interp is a subject i've been getting super enthusiastic about in 2024 since discovering @ch402's work

[Lucile Ter-Minassian @TerLucile]

takeaways: I'm learning a lot!
This project touches upon LLM-specific challenges, which I have been reading about but have not yet had hands-on experience with. I'm very excited to be part of this

[Lucile Ter-Minassian @TerLucile]

takeaways: I'm learning a lot!
This project touches upon LLM-specific challenges, which I have been reading about but have not yet had hands-on experience with. I'm very excited to be part of this

[Lucile Ter-Minassian @TerLucile]

PhD tip: there are plenty of mentorship programs for you to learn about topics outside of your lab's expertise!

I got accepted to SPAR & am now working on an AI Alignment project looking at how LLMs plan for future tokens, supervised by Nicky Pochinkov https://t.co/GanC3kY7K8

[Lucile Ter-Minassian @TerLucile]

about the project: we're investigating how LLM "plan ahead" i.e. how current-token hidden states encode data directly useful for possible future tokens. Our aim is to do so in a compute-efficient way

[Andy Arditi @andyarditi]

New research post on refusals in LLMs https://t.co/GqOo2R6lDZ

[Lucile Ter-Minassian @TerLucile]

@krvarshney congrats Kush!

[Victor Akinwande @aknvictor]

Introducing the 1st issue of the Artificial Intelligence (AI) Safety & Governance Newsletter!

https://t.co/q8J4qtJSFD

[ruchowdh.bsky.social @ruchowdh]

On the heels of Open AIs letter in safety - we need a global governance body for generative AI systems. Here’s my thoughts on how we can build one.

https://t.co/FDMpv1zzFl