Jonathan Nicholas

Very happy to see this work with Euan Prentis posted! If you’re going to CCN next week, go check out Euan’s poster on this work!

My first, first author paper, comparing the properties of memory-augmented large language models and human episodic memory, out in @cp-trendscognsci.bsky.social!

authors.elsevier.com/a/1lV174sIRv...

Here’s a quick 🧵(1/n)

Super excited to share this one!! Meta-learning sparsity and learning rate gives rise to brain-like gradients of complementary learning systems. So complementary learning systems emerge organically through behavior optimization, and it's not just two of them!!

the paper looks cool, excited to dig in!

Discovering cognitive strategies with tiny recurrent neural networks - Nature Modelling biological decision-making with tiny recurrent neural networks enables more accurate predictions of animal choices than classical cognitive models and offers insights into the underlying cog...

Thrilled to see our TinyRNN paper in @nature! We show how tiny RNNs predict choices of individual subjects accurately while staying fully interpretable. This approach can transform how we model cognitive processes in both healthy and disordered decisions. doi.org/10.1038/s415...

Neural evidence that humans reuse strategies to solve new tasks Humans can apply solutions used in past problems to new problems. In this study, the authors reveal the neural correlates of this process, known as generalization, and show that humans apply past poli...

One way to tackle a new task is to reuse solutions from the past. Check out Sam Hall-McMaster's latest finding that strategy reuse is accompanied by neural reactivation of prior solutions @plosbiology.org

Collab w/ M Tomov & @gershbrain.bsky.social

#neuroskyence
journals.plos.org/plosbiology/...

📢 now out @natrevpsychol.nature.com In a new perspective w/ @davidnagy.bsky.social & @gergoorban.bsky.social, we reconcile a glaring problem in applying rate-distortion theory as a framework for human memory, integrating empirical findings across a host of human memory research 🧵👇

I’m thrilled to announce that I will start as a presidential assistant professor in Neuroscience at the City U of Hong Kong in Jan 2026!
I have RA, PhD, and postdoc positions available! Come work with me on neural network models + experiments on human memory!
RT appreciated!
(1/5)

work with fred! can’t recommend more highly

🚨Today's Computational Psychiatry seminar at the MPC is given by Dr. Jonathan Nicholas @jonathannicholas.bsky.social
Episodic memory facilitates flexible decision making via access to individual events and their details.
2pm London UK time. DM @yanivabir.bsky.social to join.

Come see our lab's first poster at #SANS2025!!

Megan Spurney will present exciting new results revealing that rewards boost working memory accuracy for kids and teens, but less so for adults.

Friday 4/25 @ 4:15
Poster P2-A-13
Investigating Age-Related Flexibility in Cognitive Effort Allocation

The precision of hippocampal representations predicts incremental value-learning across the adult lifespan https://www.biorxiv.org/content/10.1101/2025.04.08.647815v1

See me at #COSYNE2025 poster session 2 if you want to learn more about this emerging work!

thanks, anna!

These results suggest that holding on to different details of memories can allow us to build new decision variables on-the-fly, whenever they are needed for a choice, and that perhaps one reason why we maintain detailed memories is to help us flexibly adapt to an uncertain future.

i) having detailed memories allows people to flexibly access different parts of these memories for decisions,
ii) people do this more often when it is unclear what details will be important in the future,
iii) episodic memories can be used as a “backup” if irrelevant details later become relevant.

Across four experiments and multiple replications, we tested this idea using a task that required people to encode individual events with multiple features and then to later make decisions based on these features to maximize their earnings. We found that...

Episodic memory facilitates flexible decision making via access to detailed events Our experiences contain countless details that may be important in the future, yet we rarely know which will matter and which won't. This uncertainty poses a difficult challenge for adaptive decision ...

Why do we remember so many details of our experiences even when it is unclear if we will actually ever need them?

In a new preprint, @marcelomattar.bsky.social and I asked whether this property is adaptive, because what will be relevant in the future often (usually?!) isn’t apparent.

Thank you to Ingrid Wickelgren and the team at Quanta for putting together this great piece, describing work by my lab and others on the neural representations of events

These findings are broadly consistent with predictions from work by @marcelomattar.bsky.social and recent findings from @paulbsharp.bsky.social, and we think help to reconcile some disparate results about when memories are accessed to support planning and inference

Proactive and reactive construction of memory-based preferences - Nature Communications Making new decisions requires retrieving memories, but when this occurs is unclear. The authors show that people typically access memories before a choice unless there are many related memories to con...

Excited to share a new paper with Daphna Shohamy & @nathanieldaw.bsky.social! Using fmri reactivation, we measured *when* people build preferences from memory. We found that people tend to do so before a choice, but that they also wait until choice time when options are linked to multiple memories

Lesions to Different Regions of the Frontal Cortex Have Dissociable Effects on Voluntary Persistence in Humans Deciding how long to keep waiting for uncertain future rewards is a complex problem. Previous research has shown that choosing to stop waiting results from an evaluative process that weighs the subjec...

excited to (finally!) share :)

www.jneurosci.org/content/45/6...

hello world. we have an opening for a strong theory postdoc to work in my lab on a exciting collaboration with Josh Berke and Loren Frank labs modeling and analyzing data on rat hipp-pfc-bg-da involvement in spatial maze foraging, replay, value etc. apply here: www.princeton.edu/acad-positio...

Amazing collective problem solving in ants:

I feel super lucky to have been part of this project showing how people selectively simulate actions whose value is more uncertain. To highlight one result, Haoxue shows that people preferentially simulate actions that lead to a greater number of possible future states.

If you haven’t read this yet, drop everything and do it. Thought provoking, based on fantastic & insightful examples (how much info do you need to solve Rubik’s cube?)

In PhD, while training rats on a Y-maze, I could not help thinking that the outcome was just 1 bit (left/right). I now have answers

This paper is magical.

🔔𝐍𝐄𝐖 𝐏𝐑𝐄𝐏𝐑𝐈𝐍𝐓 𝐀𝐋𝐄𝐑𝐓🔔 Beyond excited to present our new work showcasing 𝐡𝐨𝐰 𝐰𝐞 𝐜𝐚𝐧 𝐩𝐫𝐞𝐝𝐢𝐜𝐭 𝐡𝐨𝐰 𝐰𝐞𝐥𝐥 𝐲𝐨𝐮 𝐰𝐢𝐥𝐥 𝐫𝐞𝐢𝐧𝐬𝐭𝐚𝐭𝐞 𝐚 𝐧𝐞𝐰 𝐦𝐞𝐦𝐨𝐫𝐲 𝐛𝐞𝐟𝐨𝐫𝐞 𝐭𝐡𝐚𝐭 𝐦𝐞𝐦𝐨𝐫𝐲 𝐢𝐬 𝐟𝐨𝐫𝐦𝐞𝐝! Wait what? Exciting collab w/ @ptoncompmemlab.bsky.social & @chrisbaldassano.bsky.social Link: www.biorxiv.org/content/10.1... (1/11)

Check out our latest, led by @marlietandoc.bsky.social!! We find that memory for individual features of objects is rapidly distorted by the objects' category structure. Plus simulations that provide an account of how the hippocampus may contribute to this effect.