Michael Matthews

PhD student @FLAIR_Ox

[Michael Matthews @mitrma]

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/

[Clarisse Wibault @ClarisseWibault]

Mean Field Games provide a framework for modelling large populations.
ICML26 Spotlight: Introducing Recurrent Structural Policy Gradient for partially observable MFGs with common noise, benefitting from faster convergence than model-free RL, but remaining tractable, unlike DP.

[Michael Matthews @mitrma]

Work done at @FLAIR_ox with @JacksonMattT, @mcbeukman, Thomas Foster, @_aletcher, Scott Fujimoto, @cedcolas and @j_foerst.

Blog post: https://t.co/3ONUTnuJxI
GitHub: https://t.co/OihCR4rUqE
Paper: https://t.co/QW50ARXwCi
end/

[Michael Matthews @mitrma]

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/

[Michael Matthews @mitrma]

While the focus of our work is on finite goal sets, we also adapt LEO for continuous goal sets through goal quantisation, achieving competitive results with Hindsight Experience Replay in continuous control tasks.
8/

[Mikael Henaff @HenaffMikael]

Happy to share that SOL has been accepted as spotlight to ICML :) come hear about SOL in Seoul!

[Michael Matthews @mitrma]

@jsuarez I think the group would definitely be interested - after the NeurIPS deadline?
I can preload my question: NLE is unsolved +already in C +people would be very impressed if it was solved. Do you think puffer could solve it?

[Michael Matthews @mitrma]

@elliotarledge This looks really cool! How does the agent trained with pufferlib perform? + can it transfer back to original Craftax? (i.e. is this a 1-1 exact environment remake?)

[Kimbo @kimbochen]

RL researchers when they try to think of a name for their new algorithm:

[Michael Matthews @mitrma]

Very well deserved! Can't wait to see what you work on there :)

[nathan monette @nathanrmonette]

Excited to announce I’ll be joining @EugeneVinitsky at @nyutandon this autumn for a PhD!

I will be working on the intersection of game theory, reinforcement learning, and autonomous vehicles.

Thanks to everyone who helped me get to this point, especially from @FLAIR_Ox :)

[Michael Beukman @mcbeukman]

1/ As compute continues to grow and simulators continue to improve, it is becoming feasible to train RL agents for billions or trillions of timesteps. However, this is only useful if agents can continue learning over such long training horizons, which is far from given 👇

[Tim Rocktäschel @_rockt]

"The only unsaturated agentic intelligence benchmark in the world"

Excuse me? @NetHack_LE is unsaturated since 2020.

[Oscar Michel @ojmichel4]

📢Current world models aren't really modeling the world; they're modeling one agent's view of it. Partial observations ≠ world state.

Future world models will be independent of any one agent's perspective. You will be able to “drop in” any number of agents at any point in time, and a persistent world state will evolve with their interactions. Imagine a neural MMORPG server. 🧵[1/10]

[Benjamin Spiegel @superspeeg]

Why did only humans invent graphical systems like writing? 🧠✍️

In our new paper at @cogsci_soc, we explore how agents learn to communicate using a model of pictographic signification similar to human proto-writing. 🧵👇