How can #LLMs make a difference in #neuroscience and #biomedicine ? Language has more human information per bit than potentially any other form of data. https://t.co/SawHesdJFY Great collabo with @sivareddyg + @MindstateDesign. @Mila_Quebec @TheNeuro_MNI @mcgillu

Taste of what LLM-driven scientific discovery will be like: with 12 minutes of thinking, GPT-5 Pro suggested repurposing a known drug to treat an untreatable food allergy. Same exact result was found by an (at the time unpublished) peer-reviewed study. And models still improving.

The book is live today on Amazon!

Consciousness science: where are we, where are we going, and what if we get there? https://t.co/OuZbs4p3dQ

DeepSeek released an OCR model today. Their motivation is really interesting: they want to use visual modality as an efficient compression medium for textual information, and use this to solve long-context challenges in LLMs. Of course, they are using it to get more training

It’s not the thought that counts: Allostasis at the core of brain function https://t.co/ZimeMoTYnl

This is super useful! With new Formula Completions in Excel, just type "=" and Copilot proactively suggests a formula, based on the context of your sheet. Here's a great example.

How do we stop LLMs from making things up? Read Praneet Suresh's groundbreaking method to detect and eliminate AI hallucinations from within the model: https://t.co/jEX3dMz1o0 @danilobzdok @jackhtstanley

The introduction to my paper on data hacking, particularly p-hacking. For comments.

Official results are in - Gemini achieved gold-medal level in the International Mathematical Olympiad! 🏆 An advanced version was able to solve 5 out of 6 problems. Incredible progress - huge congrats to @lmthang and the team!

The biggest news of the day: John Schulman has dropped a new blog post.

🚨 BAD news for Medical AI models. MASSIVE revelations from this @Microsoft paper. 🤯 Current medical AI models may look good on standard medical benchmarks but those scores do not mean the models can handle real medical reasoning. The key point is that many models pass tests

1/ It feels surreal to announce completion of the first human trial in the development of our neurotech platform for designing mental states, from the molecular level. Human experience is now programmable. A🧵on the sequel to psychedelics & the first new "emotion in a bottle."

Now accepted at #NeurIPS 2025: Our team makes three key contributions to #large #language #models + #security: 1) Pre-trained transformer models impose semantic structure on inputs, tying them into learned conceptual webs, even if the model inputs are ambiguous or lack any

Why do LLMs hallucinate? We may have the answer! We show that LLMs imagine MORE concepts in inputs with LESS semantic structure. Very surprising! Read the thread below for more details 👇 @praneet_suresh_ @soniajoseph_ @ScimecaLuca @danilobzdok

We're thrilled to finally share what we've been working on. Our new paper gives a first-ever glimpse into the "mind" of an LLM, and we discovered something that shocked us: AIs see ghosts in the machine. 👻 What do LLMs see in the dark?

Why #language #models hallucinate? ->We have experimental data on that. We precisely locate where #hallucinations arise in LLMs + show by targeted #intervention that these internal bias features do cause bad model outputs The trick: we get LLMs drunk and they tell us the

If you think about how statistics works it’s extremely obvious why a model built on purely statistical patterns would “hallucinate”. Explanation in next tweet.

📢 job opening: software developer at Mila Quebec AI Institute & McGill As of now, my team is looking to hire a new technical staff member with abackground in a STEM area, such as computer science, engineering or physics, to support core lab activities around data analysis,

We're launching an "AI psychiatry" team as part of interpretability efforts at Anthropic!  We'll be researching phenomena like model personas, motivations, and situational awareness, and how they lead to spooky/unhinged behaviors. We're hiring - join us!

Great workshop - thanks again for inviting me to Banff.