Tom Burns

The opportunity gap in AI is more striking than ever. We talk way too much about those receiving $100M or whatever for their jobs, but not enough those asking for <$1k to present their work. For 3rd year in a row, @ml_collective is raising funds to support @DeepIndaba attendees.

Sensitivity and Sharpness of n-Simplical Attention On the topic of stabilizing training, I got unreasonably nerdsniped by the 2-simplical attention and ended up deriving the sensitivity and sharpness bounds of n-simplical attention more generally...

A gentle reminder that TMLR is a great journal that allows you to submit your papers when they are ready rather than rushing to meet conference deadlines. The review process is fast, there are no artificial acceptance rates, and you have more space to present your ideas in the…

Thrilled to collaborate on the launch of 📚 CommonPile v0.1 📚 ! Introducing the largest openly-licensed LLM pretraining corpus (8 TB), led by @kandpal_nikhil @blester125 @colinraffel. 📜: arxiv.org/pdf/2506.05209 📚🤖 Data & models: huggingface.co/common-pile 1/

甘利 俊一 栄誉研究員が「京都賞」を受賞 riken.jp/pr/news/2025/2…

Diffusion models create novel images, but they can also memorize samples from the training set. How do they blend stored features to synthesize novel patterns?  Our new work shows that diffusion models behave like Dense Associative Memory: in the low training data regime (number…

Thrilled to share our #Interspeech2025 paper: “A Silent Speech Decoding System from EEG & EMG with Heterogeneous Electrode Configurations.” We show, for the first time, that models trained on healthy data can be fine-tuned to decode the attempted speech of a totally paralyzed man

A few months ago I resigned from my tenured position at the University of Melbourne and joined Timaeus as Director of Research. Timaeus is an AI safety non-profit research organisation. [1/n]🧵

🧵 NEW POST: 'Interpreting Complexity' To what extent are we able to use differences in complexity to interpret latent structure in our models? We introduce the per-sample LLC and then do some fun stuff with it. A thread on our approach to neural network interpretability 👇

🧵 New paper! We explore sparse coding, superposition, and the Linear Representation Hypothesis (LRH) through identifiability theory, compressed sensing, and interpretability research. If neural representations intrigue you, read on! 🤓 arxiv.org/abs/2503.01824

1/ AI is accelerating. But can we ensure that AIs truly share our values and follow our goals? We argue that aligning advanced AI systems requires cracking a core scientific challenge: how data shapes AI's internal structure, and how that structure determines behavior.

I wrote a thing for @_TheTransmitter. The attack on scientific infrastructure happening in the US shows that relying on any one country is not a good option for science. We need to start supporting and building international, decentralised infrastructure for science.

Glad to see this paper out! Congrats to @anna_hedstroem and the whole team :)

SAEs were the frogs in this tweet

I am super excited to announce the call for papers for the New Frontiers in Associative Memories workshop at ICLR 2025. New architectures and algorithms, memory-augmented LLMs, energy-based models, Hopfield networks, associative memory and diffusion, and many other exciting…

Although citations are generated by many unmerited factors (geography, research area, luck, famous collaborators/supervisors/organisations, marketing/PR, etc.), I am nevertheless happy to see other researchers have found my work citable. Merry xmas and happy new year, all :)

Excited to share Dense Associative Memory through the Lens of Random Features accepted to #NeurIPS2024🎉 DenseAMs need new weights for each stored pattern–hurting scalability. Kernel methods let us add memories without adding weights! Distributed memory for DenseAMs, unlocked🔓