←
Peter Doohan41 posts
Peter Doohan avatar

Peter Doohan

@petertdoohan

PhD student @OxNeuro // Interested in how goals structure our behaviour

◷ Joined September 2022▧ 6 Photos and videos
Tim Behrens avatar
⟳ Peter Doohan retweeted
Tim Behrens@behrenstimb·Oct 26

New review about computations in PFC, led by two deep thinkers, @GabyMohamady and @kristorpjensen with contributions from many more. Discussing the implications of a bunch of exciting recent PFC findings for how we think about PFC computations.

World models are central to intelligence - but how do we actually use them to behave flexibly? We review emerging findings on how the PFC’s inner dynamics translate structured knowledge into action. osf.io/preprints/psyarxiv/j7… 🧵 1/15

Post attachment
Mohamady El-Gaby avatar
⟳ Peter Doohan retweeted
Mohamady El-Gaby@GabyMohamady·Oct 26

World models are central to intelligence - but how do we actually use them to behave flexibly? We review emerging findings on how the PFC’s inner dynamics translate structured knowledge into action. osf.io/preprints/psyarxiv/j7… 🧵 1/15

Post attachment
Tim Behrens avatar
⟳ Peter Doohan retweeted
Tim Behrens@behrenstimb·Oct 22

Super proud of many things about this work from the amazing @KrisTorpJensen. The work is one of my favourite things ever from the lab. Also super proud of Kris for choosing to send it to @elife -we get to see Kris explain every figure! elifesciences.org/articles/1…

Simone D'Ambrogio avatar
⟳ Peter Doohan retweeted
Simone D'Ambrogio@simonedambrogio·Jul 1

Thrilled to see our paper in @NatureNeuro! We combine neural networks with symbolic regression to turn how people value information into a simple, interpretable equation. We hope it's a useful recipe for making flexible models interpretable. doi.org/10.1038/s41593-026-0…

Mohamady El-Gaby avatar
⟳ Peter Doohan retweeted
Mohamady El-Gaby@GabyMohamady·Jun 11

Amazing work! Featuring one of the most convincing prefrontal silencing results I’ve ever seen.

How do brains plan actions towards goals? To get at this question we studied mice navigating complex mazes as goals changed on every trial 🧵 New work with Thomas Akam @behrenstimb @KrisTorpJensen now on BioRxiv: biorxiv.org/content/10.64898…

Tim Behrens avatar
⟳ Peter Doohan retweeted
Tim Behrens@behrenstimb·Jun 11

This is one of the coolest things I have been involved with. Truly flexible navigation in mice. Real insight into how mPFC computes the plans. Great fun to see what PFC is doing with those hippocampal theta sweeps. From the awesome @petertdoohan

How do brains plan actions towards goals? To get at this question we studied mice navigating complex mazes as goals changed on every trial 🧵 New work with Thomas Akam @behrenstimb @KrisTorpJensen now on BioRxiv: biorxiv.org/content/10.64898…

Peter Doohan avatar
Peter Doohan@petertdoohan·Jun 11

All data, analysis code & relevant docs are publicly available. Ready for anyone to start making their own discoveries 💡 See code & data availability section of manuscript :) biorxiv.org/content/10.64898…

Peter Doohan avatar
Peter Doohan@petertdoohan·Jun 11

We think these dynamics suggest a theta-paced computation for making goal-directed decisions: future positions sampled by the hippocampal formation get evaluated by their distance-to-goal in mFC to update a representation of behavioural policy 📍📏⤵️ 7/7

Peter Doohan avatar
Peter Doohan@petertdoohan·Jun 11

Both codes came together within theta oscillations, each sweeping from further to closer to the goal, but at a systematic offset 🔄 6/7

Post attachment
Peter Doohan avatar
Peter Doohan@petertdoohan·Jun 11

The second was flexible, and useful for evaluating possible futures: a population code for shortest-path distance-to-goal 5/7

Post attachment
Peter Doohan avatar
Peter Doohan@petertdoohan·Jun 11

The first resembled an efficient representation of the animals’ current behavioural policy: built from neurons tuned to conjunctions of maze place and movement direction 4/7

Post attachment
Peter Doohan avatar
Peter Doohan@petertdoohan·Jun 11

Mice used knowledge of maze structures to navigate efficiently, and in mFC we found two neural representations carrying this structured knowledge… 3/7

Peter Doohan avatar
Peter Doohan@petertdoohan·Jun 11

We thought the medial frontal cortex (mFC) might be important, and show that it is using optogenetic inhibition⚡️ 2/7