Sebastian Michelmann

@s_michelmann@indieweb.social
88 Followers
72 Following
25 Posts
Cognitive neuroscientist investigating human episodic memory, M/EEG, ECoG, and behavior. How do we reinstate vivid, information-rich memories? How do we get from A to B in memory? How does the brain represent temporally dynamic information?
Our lab is located in the heart of New York City and offers a stimulating environment. We provide access to state-of-the-art facilities, and the chance to work with a supportive and inclusive team. We strongly encourage applications from candidates of all backgrounds. (3/3)
Ideal candidates have a keen interest in human episodic memory, strong computational & analytical skills, and thrive in a collaborative environment. Experience with human electrophysiology (EEG, MEG, iEEG/ECoG) and/or neural network modeling of cognition are preferred (2/3)
Please repost! I am excited to share that I am looking for motivated postdocs who want to come and work with me at NYU (please apply here: https://apply.interfolio.com/145544)🧵⬇️ (1/3)
Apply - Interfolio

🎉 Many of you already know, but I'm thrilled to be joining NYU Psychology as an Assistant Professor this fall. I'll be looking for at least one postdoc, a lab manager, and I will be recruiting PhD students. Please retweet and stay tuned for details! Our lab will investigate Interactive Dynamics of Episodic Memory using EEG, MEG, behavioral research, and computational modeling. If this excites you, please get in touch! More info and contact details: http://s-michelmann.github.io
Interactive Dynamics of Episodic Memory

This is my personal website

Interactive Dynamics of Episodic Memory

RT @ImagingNeurosci
All NeuroImage and NeuroImage:Reports editors have resigned over the high publication fee, and are starting a new non-profit journal

https://imaging-neuroscience.org

This comes with great regret, and a huge amount of thought and discussion- please read announcement to get more details.

Imaging Neuroscience

Check out our newest preprint on memory reactivation of real-world spatial orientation during wake and sleep. Featuring @tobiasstaudigl & @b_j_griffiths

https://www.biorxiv.org/content/10.1101/2023.01.27.525854v1

RT @th_schreiner
Check out our newest preprint on memory reactivation of real-world spatial orientation during wake and sleep. Featuring @t_staudigl , @mervekutlu_ & @b_j_griffiths.

https://www.biorxiv.org/content/10.1101/2023.01.27.525854v1

The code supporting this preprint is available here: https://github.com/s-michelmann/GPT_event_segmentation
GitHub - s-michelmann/GPT_event_segmentation: code to perform event segmentation with GPT-3 and test similarity to humans

code to perform event segmentation with GPT-3 and test similarity to humans - GitHub - s-michelmann/GPT_event_segmentation: code to perform event segmentation with GPT-3 and test similarity to humans

GitHub
These findings demonstrate that GPT-3 is able to represent event structure with zero-shot prompting in a way that resembles human event perception. Going forward, this method can be used to automatically segment large text-corpora into events
Finally, we derive a continuous measure of event boundaries: the log probability of the “newline” token. This is sig. correlated with the log-probability of button-press in humans, reaching a max. correlation of r=0.372 with data from “Pieman” (left, right=”Monkey in the Middle”)