iiinova — where your digital self keeps evolving

It has been incredibly fun to work with this dream team: @hayoungsong.bsky.social, @ycleong.bsky.social, @monicarosenb.bsky.social, Marvin Chun, and Rhea Madhogarhia. I’d also like to thank @evgoldfarb.bsky.social and Nick Turk-Browne for their helpful feedback on the project! (10/10)

060

We have open-sourced the dataset (openneuro.org/datasets/ds0...) and the code to replicate this work (github.com/jinke828/soc...). (9/10)

openneuro.org

OpenNeuro

Taken together, these findings reveal that social impression updating occurs through both the ongoing neural activity during narrative processing and transient neural pattern shifts in distributed brain activity patterns at moments of insight. (8/10)

Finally, we show that distinct forms of insight selectively update complementary components of impressions: insight about the characters updates person-centered representations, whereas non-character insight shapes social-event structure. (7/10)

This motivated a whole-brain analysis. We observed significant neural pattern shifts near social insight moments in TPJ, dlPFC, dmPFC, vmPFC, precuneus, STS and sensory regions, which significantly predicted the degree of subsequent impression updating. (6/10)

Are impressions updated gradually over the course of social perception, or also through discrete moments where understanding shifts abruptly? In the rSTS, we found significant neural pattern shifts selectively near social insights, whose magnitude marginally predicted impression updates. (5/10)

Further, causal mediation modeling suggests that pre-existing impressions serve as mental priors that shape subsequent impressions and that neural synchrony in the right superior temporal sulcus (rSTS) mediates the impact of initial impression similarity on later convergence. (4/10)

How are social impressions encoded in the brain? Using IS-RSA, we found that individuals with more similar pre-existing impressions of a character exhibit higher neural synchrony when later watching them onscreen, which in turn tracks greater convergence in the subsequent impressions. (3/10)

140

In our experiment, participants viewed an audiovisual social narrative presented in a temporally scrambled order. Over 10 runs, participants watched the movie while pressing an “aha” button to indicate moments of insight and describe impressions of the characters. Data from Song et al., 2026. (2/10)

Preprint alert! 🧠 Using movie-watching fMRI and NLP, we show that prior social impressions shape how new social information is processed and updated over time, while moments of sudden insight ("aha"!) accompany transient shifts in brain activity that predict impression updates. (1/10)

Yuan Chang Leong @ycleong.bsky.social · 19/04/2026

15217

Reposted by Jin Ke

Deeply honored that our paper was recognized with the #SANS2026 Award! SANS was an important part of this paper's journey! @jadynpark.bsky.social first presented this work at SANS2023, and again in SANS2025, and we benefited greatly from the feedback and discussion!

2455

Reposted by Jin Ke

Jadyn Park @jadynpark.bsky.social · 13/10/2025

Out now in @nathumbehav.nature.com! We applied graph theoretic analyses to fMRI data of participants watching movies/listening to stories. Integration across large-scale functional networks mediates arousal-dependent enhancement of narrative memories. Open access link: rdcu.be/eKKAw

rdcu.be

Client Challenge

15719

Reposted by Jin Ke

Hayoung Song @hayoungsong.bsky.social · 05/09/2025

How does the brain🧠 make causal inferences and use memories to understand narratives🎬? We built an RNN🤖 with key-value episodic memory that learns causal relationships between events and retrieves memories like humans do! Preprint www.biorxiv.org/content/10.1... w/ @qlu.bsky.social, Tan Nguyen &👇

A neural network with episodic memory learns causal relationships between narrative events

Humans reflect on past memories to make sense of an ongoing event. Past work has shown that people retrieve causally related past events during comprehension, but the exact process by which this causa...

39529

I'm also deeply grateful to those who offered insightful feedback along the way: the BLRB community at UChicago, joint CogNeuro meeting at Yale, @esfinn.bsky.social 's lab and many others. I’d like to thank Emma Megla and Wilma Bainbridge for sharing the aphantasia data! (10/10)

HUGE thank you to everyone whose incredible efforts over the past three years made this project possible! @tchamberlain.bsky.social @hayoungsong.bsky.social @annacorriveau.bsky.social @zz112.bsky.social, Taysha Martinez, Laura Sams, Marvin Chun, @ycleong.bsky.social @monicarosenb.bsky.social (9/10)

150

Together, we found that ongoing thoughts at rest are reflected in brain dynamics and these network patterns predict everyday cognition and experiences. Our work underscores the crucial role of subjective in-scanner experiences in understanding functional brain organization and behavior. (8/10)

Neuromarkers of these thoughts further generalized to HCP data (N=908), where decoded thought patterns predicted positive vs. negative trait-level individual differences measures. This suggests that links between rsFC and behavior might in part reflect differences in ongoing thoughts. (7/10)

Moreover, the model predicting whether people are thinking in the form of images distinguished an aphantasic individual—who lacks visual imagery—from their otherwise identical twin. Data from academic.oup.com/cercor/artic.... (6/10)

150

Thought models generalized beyond self-report, predicting non-introspective markers, such as pupil size, linguistic sentiment of speech and the strength of a sustained attention network (Rosenberg et al., 2016, 2020). (5/10)

How are these thoughts related to resting-state functional connectivity (rsFC) patterns? We found that similarity in ongoing thoughts tracks similarity in rsFC patterns within and across individuals, and that both thought ratings and topics could be reliably decoded from rsFC (4/10)

We observed a remarkable idiosyncrasy in ongoing thoughts between individuals and over time, both in terms of self-reported ratings as well as the content and topics of thoughts. (3/10)

121

In our “annotated rest” task, 60 individuals rested, and verbally described and rated their ongoing thoughts after each 30-sec rest period. (2/10)

151

New preprint! 🧠 Our mind wanders at rest. By periodically probing ongoing thoughts during resting-state fMRI, we show these thoughts are reflected in brain network dynamics and contribute to pervasive links between functional brain architecture and everyday behavior (1/10). doi.org/10.1101/2025...

Ongoing thoughts at rest reflect functional brain organization and behavior

Resting-state functional connectivity (rsFC)-brain connectivity observed when people rest with no external tasks-predicts individual differences in behavior. Yet, rest is not idle; it involves streams...

46923

Reposted by Jin Ke

Hayoung Song @hayoungsong.bsky.social · 12/08/2025

Preprint⭐ Our attention changes over time and differs across contexts—which is reflected in the brain🧠 Fitting a dynamical systems model to fMRI data, we find that the geometry of neural dynamics along the attractor landscape reflects such changes in attention! www.biorxiv.org/content/10.1...

Geometry of neural dynamics along the cortical attractor landscape reflects changes in attention

The brain is a complex dynamical system whose activity reflects changes in internal states, such as attention. While prior work has shown that large-scale brain activity reflects attention, the mechan...

16021

Reposted by Jin Ke

Qihong (Q) Lu @qlu.bsky.social · 08/05/2025

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)

1413040

Reposted by Jin Ke

Yuan Chang Leong @ycleong.bsky.social · 27/04/2025

Feeling fortunate that #SANS2025 was in Chicago, and so much of the lab was able to be part of the meeting! It's crazy how much the lab has grown over the past 3y10m, and I'm so proud of the work we are doing together! Happy that we could host the lab, alums (and surprise guests)! #CASNL@SANS

1456

Jin Ke @jinke.bsky.social · 24/04/2025

To learn more about this dataset and the neural dynamics of narrative insight, check out our recent work (preprint below) led by the amazing @hayoungsong.bsky.social and chat with her on Saturday 1:50 PM - 3:00 PM! Poster ID: P3-B-30. www.biorxiv.org/content/10.1...

021

Jin Ke @jinke.bsky.social · 24/04/2025

Curious how the human brain updates social impression in a naturalistic setting? We scanned participants watching This Is Us, and found that sudden neural pattern shifts at insight moments of comprehension reflect impression updating. Come and chat Friday 4:15–5:15pm at #SANS2025, Poster P2-G-69.

1223

Reposted by Jin Ke

Erica Busch @elbusch.bsky.social · 04/04/2025

New preprint! Excited to share our latest work “Accelerated learning of a noninvasive human brain-computer interface via manifold geometry” ft. outstanding former undergraduate Chandra Fincke, @glajoie.bsky.social, @krishnaswamylab.bsky.social, and @wutsaiyale.bsky.social's Nick Turk-Browne 1/8

doi.org

Accelerated learning of a noninvasive human brain-computer interface via manifold geometry

Brain-computer interfaces (BCIs) promise to restore and enhance a wide range of human capabilities. However, a barrier to the adoption of BCIs is how long it can take users to learn to control them. W...

26619

Many thanks to Janice Chen, @lukejchang.bsky.social and @asieh.bsky.social for open sourcing the movie datasets! Also wanted to give a shoutout to UChicago MRIRC for helping us collect the North by Northwest data. (9/9)

030

GitHub - jinke828/AffectPrediction

We have made our model and analysis scripts publicly available to facilitate its use by other researchers in decoding moment-to-moment emotional arousal in novel datasets, providing a new tool to probe affective experience using fMRI. (8/9) github.com/jinke828/Aff...

github.com

Contribute to jinke828/AffectPrediction development by creating an account on GitHub.

110

In conclusion, our findings reveal a generalizable representation of emotional arousal embedded in patterns of dynamic functional connectivity, suggesting a common underlying neural signature of emotional arousal across individuals and situational contexts. (7/9)

100

In contrast, using the same computational modeling approach, we were unable to find a generalizable neural representation of valence in functional connectivity. Null results are inherently difficult to interpret, but we offer several possible speculations in our paper. (6/9)

100

The network generalized to two additional, novel movies, where model-predicted arousal time courses corresponded with the plot of each movie, suggesting a methodological tool for researchers who wish to obtain continuous measures of arousal without having to collect additional human ratings. (5/9)

100

This generalizable arousal network is encoded in interactions between multiple large-scale functional networks including default mode network, dorsal attention network, ventral attention network and frontoparietal network. (4/9)

100

We observed robust out-of-sample generalizability of the arousal models across movie datasets that were distinct in low-level features, characters, narratives, and genre, suggesting a situation-general neural representation of arousal. (3/9)

100

One possibility is that there are generalizable neural patterns associated with valence or arousal across contexts and individuals. We utilized open movie-watching fMRI datasets and built predictive models of moment-to-moment valence and arousal from functional correlations in brain activity. (2/9)

100

From the excitement of reuniting with a long-lost friend to the anger when being treated unfairly, our daily lives are colored by diverse affective experiences. Despite the remarkable diversity, is there an intrinsic similarity in how the brain represents these experiences? (1/9)