←
David Abel634 posts
David Abel avatar

David Abel

@dabelcs

Scientist @GoogleDeepMind and Fellow @EdinburghUni | RL, philosophy, agency, foundations, alignment | office hours: tinyurl.com/dabel-mtg

◷ Joined January 2018▧ 30 Photos and videos
David Abel avatar
David Abel@dabelcs·Sep 29

RLDM is returning in 2027!

📣 Save the date! RLDM 2027, the Multi-disciplinary Conference on Reinforcement Learning and Decision Making, is coming to Paris, July 6-9, 2027 🎉 Stay tuned for submission deadlines and more updates on our website rldm.org 👀

Post attachment
RLDM avatar
⟳ David Abel retweeted
RLDM@RLDMParis2027·Sep 29

📣 Save the date! RLDM 2027, the Multi-disciplinary Conference on Reinforcement Learning and Decision Making, is coming to Paris, July 6-9, 2027 🎉 Stay tuned for submission deadlines and more updates on our website rldm.org 👀

Post attachment
Tom Silver avatar
⟳ David Abel retweeted
Tom Silver@tomssilver·Jul 5

This week's #PaperILike is "General agents contain world models" (Richens et al., ICML 2025). Brought to my attn by @dabelcs 's beautiful philosophy-laden talk at ICAPS. The explicit construction of a transition model from a policy is cool (Alg 1). PDF: arxiv.org/abs/2506.01622

Alistair Letcher avatar
⟳ David Abel retweeted
Alistair Letcher@_aletcher·Jun 23

Model-free agents learn to maximise reward without modelling the environment. Right? In recent work, we challenge this narrative by proving that agents, trained on a sufficiently rich set of goals, encode a unique and accurate world model in their value functions. 1/

Post attachment
Geoffrey Irving avatar
⟳ David Abel retweeted
Geoffrey Irving@geoffreyirving·Jun 10

We are starting a new, nonprofit alignment organization, ⊢ Sequent Research, bringing together researchers previously on UK AISI’s Alignment Team, Timaeus, and elsewhere to research how to align superintelligence. We are hiring! 🧵

Post attachment
Brian Christian avatar
⟳ David Abel retweeted
Brian Christian@brianchristian·Jun 3

Just published in @PNASNews, we resolve a 50-year-old riddle from Richard Feynman's handwritten notes, prove and generalize it, and run a large-scale human study to reveal near-optimal heuristics in sequential decision problems: doi.org/10.1073/pnas.2509612…

alphaXiv avatar
⟳ David Abel retweeted
alphaXiv@askalphaxiv·May 25

"Imperfect World Models are Exploitable" World models can look accurate, but still rank policies incorrectly, saying policy A is better than policy B when the real environment says the reverse. This paper formalizes that failure as model exploitation and proves it is basically unavoidable for any nontrivial, nonequivalent world model on broad policy sets. It also connects this to reward hacking and derives a safe horizon showing how model error compounds with planning depth.

Post attachment
Finding The Frame Workshop avatar
⟳ David Abel retweeted
Finding The Frame Workshop@RLFrameWorkshop·May 13

Reminder that the deadline for the Finding the Frame workshop @RL_Conference is coming up on May 22 AoE ⏰ We're excited to read your submissions reflecting on the philosophy, practice, and formalisms of reinforcement learning! 🔗 More details: sites.google.com/view/findin…

Raul Steleac avatar
⟳ David Abel retweeted
Raul Steleac@steleac·Apr 20

Really excited to present our recent work at #ICLR2026 this week! We discover highly coordinated joint behaviours and integrate them into the skill sets of MARL agents, accelerating the search for effective joint strategies in downstream tasks.🧵 Paper: raulsteleac.github.io/iaro

Post attachment