Jonny Cook

Can an LLM be programmed? In our new preprint, we show that LLMs can learn to evaluate programs for a range of inputs by being trained on the program source code alone – a phenomenon we call Programming by Backprop (PBB). 🧵⬇️

Since R1 there has been a lot of chatter 💬 on post-training LLMs with RL. Is RL only sharpening the distribution over correct responses sampled by the pretrained LLM OR is it exploring and discovering new strategies 🤔? Find answers in our latest post ⬇️ tinyurl.com/rlshadis

1/ 🕵️ Algorithm discovery could lead to huge AI breakthroughs! But what is the best way to learn or discover new algorithms? I'm so excited to share our brand new @rl_conference paper which takes a step towards answering this! 🧵

Unlock the Hidden Diversity in Your Language Model. In our new paper, Intent Factored Generation (IFG), we propose an inference time method to increase the diversity of generations from LLMs. IFG leads to improvements in searching for solutions to maths and code problems. (1/6)

General relativity 🤝 neural fields This simulation of a black hole is coming from our neural networks 🚀 We introduce Einstein Fields, a compact NN representation for 4D numerical relativity. EinFields are designed to handle the tensorial properties of GR and its derivatives.

🚨 Did you know that small-batch vanilla SGD without momentum (i.e. the first optimizer you learn about in intro ML) is virtually as fast as AdamW for LLM pretraining on a per-FLOP basis? 📜 1/n

How can we unlock generalized reasoning? ⚡️Introducing Energy-Based Transformers (EBTs), an approach that out-scales (feed-forward) transformers and unlocks generalized reasoning/thinking on any modality/problem without rewards. TLDR: - EBTs are the first model to outscale the…

1/n Multi-token prediction boosts LLMs (DeepSeek-V3), tackling key limitations of the next-token setup: • Short-term focus • Struggles with long-range decisions • Weaker supervision Prior methods add complexity (extra layers) 🔑 Our fix? Register tokens—elegant and powerful

Antiviral therapy design is myopic 🦠🙈 optimised only for the current strain. That's why you need a different Flu vaccine every year! Our #ICML2025 paper ADIOS proposes "shaper therapies" that steer viral evolution in our favour & remain effective. Work done @FLAIR_Ox 🧵👇

Theory of Mind (ToM) is crucial for next gen LLM Agents, yet current benchmarks suffer from multiple shortcomings. Enter 💽 Decrypto, an interactive benchmark for multi-agent reasoning and ToM in LLMs! Work done with @TimonWilli & @j_foerst at @AIatMeta & @FLAIR_Ox 🧵👇

The way programs are represented in training data can have strong effects on generalization & reasoning -- really great (and truly scientific) work

How can we bypass the need for online hyper-parameter tuning in offline RL? @FLAIR_Ox is introducing two fully offline algorithms: SOReL, for accurate offline regret approximation, and TOReL, for offline hyper-parameter tuning! arxiv.org/html/2505.2244…

LLMs can be programmed by backprop 🔎 In our new preprint, we show they can act as fuzzy program interpreters and databases. After being ‘programmed’ with next-token prediction, they can retrieve, evaluate, and even *compose* programs at test time, without seeing I/O examples.

🧵 Check out our latest preprint: "Programming by Backprop". What if LLMs could internalize algorithms just by reading code, with no input-output examples? This could reshape how we train models to reason algorithmically. Let's dive into our findings 👇