Twitxy Saved

Filip Trhlík

@FilipTrhlik

ML & NLP PhD at Cambridge | Just vibing with LLMs
Joined January 2021
41 Following 32 Followers 11 Posts

Latest posts 2

Filip Trhlík @FilipTrhlik
Beyond excited that our work with UPenn NLP on examining the robustness of LLM detection is finally out.

#ACL2024
Liam Dugan @LiamDugan_ Jun 20, 2024
🚨New Paper🚨: Are AI text detectors *really* as good as they claim? (#ACL2024)

We release RAID—The largest & most challenging detection benchmark with 6M+ outputs from 11 LLMs, 8 domains, 4 decoding strategies, and 11 adv attacks

arxiv.org/abs/2405.07940
github.com/liamdugan/raid
View quoted post on X