Théo Lasnier (@Blyzi1) profile picture

PhD Student | @inria_paris ALMAnaCH | LLM Interpretability

Joined December 2018
Photo by Théo Lasnier (@Blyzi1)
Théo Lasnier retweeted
🚨Our new paper, TrustMI, is out! Watch Théo's video 👇 We study how LLMs represent trust in their users—and how to steer it. TL;DR: more trust often meant less safety; lower trust reduced harm in our tests. 🧵1/X
Did you ever wonder if your AI model trusted you ? We just release a new paper TrustMI where we finds a linear trust direction in 6 LLMs. 📄 arxiv.org/abs/2610.06064 🤗 huggingface.co/TrustMI
1
2
10
410
Did you ever wonder if your AI model trusted you ? We just release a new paper TrustMI where we finds a linear trust direction in 6 LLMs. 📄 arxiv.org/abs/2610.06064 🤗 huggingface.co/TrustMI
3
5
2
14
1,450
Huge thanks to my co-authors @froger_romain, @maxencelsb and @zehavoc !!!
3
100