Siddharth Mishra-Sharma

icon indicating a website link smsharma.github.io icon indicating an email [email protected]

icon company MIT, IAIFI icon location Cambridge, MA icon location Physicist and machine learning researcher. IAIFI Fellow at MIT/Harvard. Probabilistic modeling/inference and AI for science/physics.

Results 7 comments of Siddharth Mishra-Sharma

Feature request: correlation of vector quantities

Thank you for the quick response! Ideally, the correlation would be between vectors projected onto the plane perpendicular to the line-of-sight from the Earth's position. This is because velocities are...

Feature request: correlation of vector quantities

Sorry---I definitely misunderstood your question! Yes, accumulating the quantities you mentioned (product of velocities along the line-of-sight and perpendicular to it) would be ideal for our application.

Normalizing flow example

Thanks, @awni, no hurry obviously!

Normalizing flow example

Closing this as the examples repo seems primarily focused on LLM-style examples.

Normalizing flow example

@awni Ah, happy to reopen! Just thought I may have misunderstood the scope of the repo. Thanks!

Normalizing flow example

Thanks a lot @awni, everything looks good! Re: class structure, agreed it doesn't add very much in particular for `distributions.py`, so I've gone ahead and simplified it by removing the...

Normalizing flow example

Thank you @awni! Been fun playing around with MLX, and I think it could be nice for general Bayesian inference in the scientific context, for which this was a little...