Pinned Tweet
🌹 Today we're releasing Unifloral, our new library for Offline Reinforcement Learning!
We make research easy:
⚛️ Single-file
🤏 Minimal
⚡️ End-to-end Jax
Best of all, we unify prior methods into one algorithm - a single hyperparameter space for research! ⤵️
Matthew Jackson retweeted
I think I have found a proof of the Forsythe Conjecture for s=2 using Codex, open since 1968.
A detailed check by a leading expert, who has called the proof "highly likely to be correct," is in progress.
Matthew Jackson retweeted
Hindsight Experience Replay has become the ubiquitous method for goal-conditioned reinforcement learning, but leaves open the question of which goal to relabel with.
In this work, accepted at ICML, we propose instead simply Learning Everything All at Once (LEO).
1/
Matthew Jackson retweeted
What if a world model could render not an imagined place, but the actual city?
We introduce Seoul World Model, the first world simulation model grounded in a real-world metropolis.
TL;DR: We made a world model RAG over millions of street-views.
proj: seoul-world-model.github.io/
Matthew Jackson retweeted
PERSIST is a world model that ditches pixel-based histories for a 3D world state. Instead of searching through an ever-growing sequence of past pixel observations, PERSIST retrieves spatial information from a dynamically evolving 3D representation.
This change improves the spatial memory, 3D consistency, and long-horizon stability of the model, enabling interactive experiences within coherent and evolving 3D worlds.
#MachineLearning #WorldModels #GenerativeAI #3DComputerVision #ComputerVision #Genie3 #AI
Matthew Jackson retweeted
We tested this loop in Genie 3—a foundation world model, creating photorealistic environments on the fly. 🧞♂️
When SIMA 2 was trained only on “Urban” worlds from Genie 3, it still significantly improved its capabilities in unseen "Natural" environments.
We are moving from static datasets to infinite, procedurally generated training grounds.
[2/N]
Matthew Jackson retweeted
🌹 has landed.
Presentation at 10:20AM Exhibit Hall FGH
Poster Exhibit 11 AM Exhibit Hall CDE
Come talk to me, @JacksonMattT @JarekLiesen about inventing new RL Algos and getting some **stickers**.
#NeurIPS2025 #offlineRL
#unifloral
Matthew Jackson retweeted
My Oxford lab (@FLAIR_Ox ) is hiring Phd students! If you are thinking of doing a Phd in blue-sky and -sort of crazy ambitious- ML and have a technically strong background and love to work with others, please consider all options for joining us:
1) Direct entry - deadline is the 1st of Dec AOE (ox.ac.uk/admissions/graduate…)
2) AIMS CDT (ox.ac.uk/admissions/graduate…) deadline on 27th of Jan 2026 AOE
3) EIT CDT (ox.ac.uk/admissions/graduate…) deadline on the 7th of Jan 2026 AOE
Student funding is a real constraint / concern in the UK (especially for overseas students) and by applying for these three programs you can maximize your chances of ending up in a very very special place.
Matthew Jackson retweeted
SIMA 2 is our most capable AI agent for virtual 3D worlds. 👾🌐
Powered by Gemini, it goes beyond following basic instructions to think, understand, and take actions in interactive environments – meaning you can talk to it through text, voice, or even images. Here’s how 🧵
Matthew Jackson retweeted
We release a paper on Jasmine, our production-ready JAX-based codebase for world modeling from unlabeled videos!
Matthew Jackson retweeted
🎮 How can agents learn to generalize from limited offline data?
We introduce iMac (Imagined Autocurricula) - training agents entirely in world models with emergent curricula!
Matthew Jackson retweeted
Super glad all the hard work has been recognised with an oral by the community. Great job team!
Now we need to do project V for online RL as well ..
Unifloral has been accepted as an Oral at NeurIPS 2025!
Immensely grateful to my @FLAIR_Ox co-authors @uljadb99 and @JarekLiesen for pouring months of effort into this project.
There’s a ton of low-hanging fruit in offline RL… If you’re looking for a project, check it out!
Unifloral has been accepted as an Oral at NeurIPS 2025!
Immensely grateful to my @FLAIR_Ox co-authors @uljadb99 and @JarekLiesen for pouring months of effort into this project.
There’s a ton of low-hanging fruit in offline RL… If you’re looking for a project, check it out!
Matthew Jackson retweeted
The field of robotics is undergoing a historic revolution right now. I’ve spent the last year thinking about how to mentally model the breakneck progress in robotics + AI. With the help of mascots like “The AGI Bro”, we can try to sift through the noise 🧵
Matthew Jackson retweeted
Was amazing to present my first paper at @RL_Conference !! Really awesome to meet new folks from the community :)
Awesome to see @prob_doom turn our Genie implementation into a hyper-optimized world modeling framework! 🚀
Highly recommend anyone in the field check it out - it’s an exciting time to be modeling worlds 🌐
Mine and @TimonWilli's Genie 1 implementation: github.com/FLAIROx/jafar
Excellent advice!
Watching @jparkerholder and team put this together has definitely been a career highlight 😄🧞
Matthew Jackson retweeted
Genie 3 feels like a watershed moment for world models 🌐: we can now generate multi-minute, real-time interactive simulations of any imaginable world. This could be the key missing piece for embodied AGI… and it can also create beautiful beaches with my dog, playable real time