It was a pleasure to host Nishant Balepur at Imperial College London this week!

Congrats to Mingxue (Mercy) Xu for this new work accepted to SLLM at ICLR 2025!

Many congratulations to Matthieu Meeus Marek Rei and coauthors for the best paper award at SaTML Conference 🎉

Can't wait to see everyone at ICLR and NAACL! Check out some of our awesome papers. Come and say hi, we'd love to have a chat :)

You can find me at ICLR 2025 in Singapore, where I’ll be presenting my work “Latent Representation Encoding and Multimodal Biomarkers for Post-Stroke Speech Assessment” on Sunday 27th, Hall 4 #6 and on Monday 28th, Peridot 201&206 🥳 ICLR 2026 Imperial NLP Imperial EEE

I’ll be presenting Meta-Reasoning Improves Tool Use in Large Language Models at #NAACL25 tomorrow Thursday May 1st from 2 until 3.30pm in Hall 3! Come check it out and have a friendly chat if you’re interested in LLM reasoning and tools 🙂 #NAACL

Congratulations to the authors of IrokoBench funded by Lacuna Fund for winning one of the Outstanding Paper Award at #NAACL2025 with Happy B. Kasumba Shamsuddeen Hassan Muhammad, PhD Israel Abebe | እስራኤል አበበ Rooweither Mabuya Andie Bukula

Happening Thursday at Imperial College London London. imperial.ac.uk/events/193952/…

Thrilled to share our new preprint on Reinforcement Learning for Reverse Engineering (RLRE) 🚀 We demonstrate that human preferences can be reverse engineered effectively by pipelining LLMs to optimise upstream preambles via reinforcement learning 🧵⬇️

🙌Happy to share our paper, “DiffuseDef: Improved Robustness to Adversarial Attacks via Iterative Denoising” is accepted to #ACL2025! Great thanks to my co-authors Huichi Zhou, Marek Rei, Lucia Specia! Arxiv: arxiv.org/abs/2407.00248 GitHub: github.com/Nickeilf/Diffu…

We have a new paper up on arXiv! 🥳🪇 The paper tries to improve the robustness of closed-source LLMs fine-tuned on NLI, assuming a realistic training budget of 10k training examples. Here's a 60 second rundown of what we found!

If you're interested in NLI or model robustness, this is a nice and fun paper - check it out 🙂 x.com/_joestacey_/st…

If anyone has done any work improving the robustness of NLI models and we didn't cite you in our appendix, please share a link to your work - I would love to include it Our appendix related work on NLI robustness is a bit of a monster 😅🧌 x.com/_joestacey_/st…

🚨 New preprint alert – accepted at #ACL2025 Findings!🚨

How good can privacy attacks against LLM pretraining get if you assume a very strong attacker? Check it out in our preprint ⬇️

(1/9) LLMs can regurgitate memorized training data when prompted adversarially. But what if you *only* have access to synthetic data generated by an LLM? In our ICML Conference paper, we audit how much information synthetic data leaks about its private training data 🐦🌬️

Welcome to Imperial College London Oana-Maria Camburu 🎉

Check out our recent work on prompt injection attacks! Tl;DR: aligned LLMs show to defend against prompt injection; yet with a strong attacker (GCG on steroids), we find that successful attacks (almost) always exist, but are just harder to find. arxiv.org/pdf/2505.15738