John J. Vastola
@johnjvastola
Followers
306
Following
228
Statuses
78
Professional Question Asker @ Harvard | brains, biology, physics, ML, math, etc
Boston, MA
Joined August 2022
@aran_nayebi Agree! Should’ve read the slides first, my bad. They look beautiful and incredibly thoughtful about the philosophy underlying looking for task-relevant reps, so kudos!
0
0
1
@GorinGennady Worth adding that even neural network methods don’t make “no assumptions”, but rather have specific (architecture- and training-details- and implementation-dependent) inductive biases that control how training data get generalized! Some theory has been done to characterize these.
0
0
2
RT @JieyuZheng3: Huge thanks to @johnjvastola for the kind invitation and intro! 🙏 And to the @harvardmed neuroscience audience for your qu…
0
1
0
RT @rlbrainseminar: Next meeting Thurs, Nov 14 at 12:15 PM in WAB 236! Speaker is Jieyu Zheng (@JieyuZheng3), a Caltech PhD student in the…
0
1
0
RT @rlbrainseminar: Next meeting is Thurs, Sep 12 at 12:15 PM in WAB 236! The speaker is Zhang-Wei Hong (@ZhangWeiHong9) from the Improbabl…
0
1
0
RT @rlbrainseminar: The Fall 2024 iteration of the RL and the Brain seminar at Harvard Medical School begins soon---this Thursday, Sep 12!…
0
1
0
The ML from Scratch seminar at HMS returns next Tuesday (Feb 20) at 5 PM in WAB 236. The first session will be on generalized linear models (GLMs). @kiahhardcastle has graciously agreed to be there to answer your GLM-related questions! For more info:
0
0
4
RT @rlbrainseminar: The seminar returns on Thurs, Jan 25 at 1:30 PM in WAB 236! Our first speaker is Ryan Badman (@RyanPaulBadman1) from th…
0
6
0
RT @WangBinxu: Xprint alert! ⏰ With @johnjvastola, we present "The Hidden Linear Structure in Score-Based Models and its Application" in th…
0
12
0
Just finished presenting work on causal inference and navigation at #SfN2023! Based on an ongoing collaboration between the Angelaki Lab (esp. Valentina Vencato and @JP__NOEL), the DeAngelis Lab, and the lab of @jdrugowitsch.
0
2
24
Work done with @GorinGennady and @lpachter. Dr. Gorin led this work, doing most of the analysis/writing/math. I've been blown away by how carefully and deeply he's thought about the foundational issues we study here. He recently defended and is job hunting. Please hire him!
0
0
7