We also did some digging to get at the specific networks that seem to exhibit this mode switching process most strongly (DMN!) or weakly (primary sensory cortex!) and also used some cool @neurosynth analyses to help us interpret the results.

When we are engrossed in a challenging or highly engaging task, our brain switches to a mode that prioritizes its ability to distinguish one moment’s activity patterns from other moments’ patterns.

Our study’s core finding is that the brain reconfigures itself from moment to moment to maintain a tradeoff between computational complexity and robustness to noise, according to ongoing cognitive demands.

For context, let's appreciate for a moment that our brains are responsible for every human thought, story, relationship, artwork, and scientific discovery EVER. Meanwhile, they're packed into a volume of under 1.5L and they run on less power than a standard incandescent 💡!

Chatify can help with explaining code, debugging, brainstorming project ideas, test your understanding, and more.

If you've installed/loaded the extension, your notebook now has AI superpowers 🦸🦹🤖! You can add the "%%explain" magic command to the top of any code cell to interact with Chatify:

We found that *covert* changes to the focus of category- and location-based volitional attention affect memory encoding at different timescales and likely through different mechanisms!

MAJOR ANNOUNCEMENT!! Effective immediately, all lab activities will now be performed by LLMs fine-tuned on our papers. Lab members will serve only as a physical interface to the external world.

Wanna participate in the upcoming (free!) spring 2023 offering of Intro to Programming for Psych Scientists? Check out our course materials (github.com/ContextLab/cs-for…) and join our Discord workspace to participate in discussions, ask questions, etc.: discord.gg/AdDgJd7cbM

We did some interesting exploratory things too, using a combination of @neurosynth and ChatGPT to help understand what different patterns we found might “mean” from a functional perspective.

We also zoomed in on specific networks. Activity from higher-order brain areas was generally more informative than from lower-order areas, but we didn’t see any obvious differences in compressibility across networks.

As the story progresses, these patterns get stronger! After listening for a while, activity evoked by the intact story or coarse scrambling becomes even *more* informative and compressible, whereas finely scrambled/rest activity becomes *less* informative and compressible.

Brain activity from people listening to the unscrambled story was both more informative overall *and* more compressible than activity during scrambled listening or during rest.

The core idea is that more “informative” brain patterns should yield higher classification accuracy. More “compressible” brain patterns should yield higher accuracy for a fixed number of features. We’re interested in tradeoffs between the two, under different circumstances.

This could be used in a huge variety of scenarios! We use it to model student-teacher interactions, but it could also be used to model conversations between friends, business partners, romantic partners, etc.