Rupesh Srivastava
@rupspace.bsky.social
AI Researcher. (Co)developed Highway Networks, Upside-Down RL, Bayesian Flow Networks, EvoTorch
📜 Learning is compression
https://rupeshks.cc/
📜 Learning is compression
https://rupeshks.cc/
Reposted by Rupesh Srivastava
*Weighted Skip Connections are Not Harmful for Deep Nets*
by @rupspace.bsky.social
Cool blog post "in defense" of weighted variants of ResNets (aka HighwayNets) - as a follow up to a previous post by @giffmana.ai.
rupeshks.cc/blog/skip.html
by @rupspace.bsky.social
Cool blog post "in defense" of weighted variants of ResNets (aka HighwayNets) - as a follow up to a previous post by @giffmana.ai.
rupeshks.cc/blog/skip.html
Weighted Skip Connections are Not Harmful for Deep Nets
Give Gates a Chance
rupeshks.cc
February 18, 2025 at 9:49 AM
*Weighted Skip Connections are Not Harmful for Deep Nets*
by @rupspace.bsky.social
Cool blog post "in defense" of weighted variants of ResNets (aka HighwayNets) - as a follow up to a previous post by @giffmana.ai.
rupeshks.cc/blog/skip.html
by @rupspace.bsky.social
Cool blog post "in defense" of weighted variants of ResNets (aka HighwayNets) - as a follow up to a previous post by @giffmana.ai.
rupeshks.cc/blog/skip.html
These checks are very important and useful. Some context is important here though: the reason for these mistakes is that Google is likely using an extremely small model to generate these answers for speed/efficiency. GPT-4o, Gemini Advanced, and even Gemini 1.5 Flash easily answer all correctly.
January 17, 2025 at 10:46 PM
These checks are very important and useful. Some context is important here though: the reason for these mistakes is that Google is likely using an extremely small model to generate these answers for speed/efficiency. GPT-4o, Gemini Advanced, and even Gemini 1.5 Flash easily answer all correctly.
Wrote a post about Highway networks, ResNets and subtleties of architecture comparisons:
rupeshks.cc/blog/skip.html
rupeshks.cc/blog/skip.html
Weighted Skip Connections are Not Harmful for Deep Nets
Give Gates a Chance
rupeshks.cc
January 11, 2025 at 1:00 AM
Wrote a post about Highway networks, ResNets and subtleties of architecture comparisons:
rupeshks.cc/blog/skip.html
rupeshks.cc/blog/skip.html
Reposted by Rupesh Srivastava
Getting myself set up here. I found the Sky Follower Bridge Chrome plugin pretty helpful (thanks @kawamataryo.bsky.social!)
chromewebstore.google.com/detail/sky-f...
chromewebstore.google.com/detail/sky-f...
Sky Follower Bridge - Chrome Web Store
Easily transfer your following users and list members from X to Bluesky.
chromewebstore.google.com
January 5, 2025 at 10:51 PM
Getting myself set up here. I found the Sky Follower Bridge Chrome plugin pretty helpful (thanks @kawamataryo.bsky.social!)
chromewebstore.google.com/detail/sky-f...
chromewebstore.google.com/detail/sky-f...
December 6, 2024 at 10:58 PM
❤️❤️❤️
Excited to reveal Genie 2, our most capable foundation world model that, given a single prompt image, can generate an endless variety of action-controllable, playable 3D worlds. Fantastic cross-team effort by the Open-Endedness Team and many other teams at Google DeepMind! 🧞
Introducing 🧞Genie 2 🧞 - our most capable large-scale foundation world model, which can generate a diverse array of consistent worlds, playable for up to a minute. We believe Genie 2 could unlock the next wave of capabilities for embodied agents 🧠.
December 4, 2024 at 8:12 PM
❤️❤️❤️
Reposted by Rupesh Srivastava
So I'm not here because it's a left-leaning space or anything like that. I'm here because helping prop up a propaganda machine feels really distasteful to me
November 30, 2024 at 2:09 PM
So I'm not here because it's a left-leaning space or anything like that. I'm here because helping prop up a propaganda machine feels really distasteful to me
Regarding creating and sharing BlueSky datasets: I feel like we're talking past each other.
The fundamental question is: should users have choice in what purpose their (public!) posts are used for?
@bsky.app needs to think through what their answer is. (1/3)
The fundamental question is: should users have choice in what purpose their (public!) posts are used for?
@bsky.app needs to think through what their answer is. (1/3)
November 30, 2024 at 12:19 AM
Regarding creating and sharing BlueSky datasets: I feel like we're talking past each other.
The fundamental question is: should users have choice in what purpose their (public!) posts are used for?
@bsky.app needs to think through what their answer is. (1/3)
The fundamental question is: should users have choice in what purpose their (public!) posts are used for?
@bsky.app needs to think through what their answer is. (1/3)
Reposted by Rupesh Srivastava
Posting a call for help: does anyone know of a good way to simultaneously treat both POTS and Ménière’s disease? Please contact me if you’re either a clinician with experience doing this or a patient who has found a good solution. Context in thread
November 24, 2024 at 4:34 PM
Posting a call for help: does anyone know of a good way to simultaneously treat both POTS and Ménière’s disease? Please contact me if you’re either a clinician with experience doing this or a patient who has found a good solution. Context in thread
Reposted by Rupesh Srivastava
the remarkable success of the Google brain (and OpenAI) resident programs is an indication to me that smart, hardworking people can do more than you expect
November 25, 2024 at 4:09 PM
the remarkable success of the Google brain (and OpenAI) resident programs is an indication to me that smart, hardworking people can do more than you expect
Reposted by Rupesh Srivastava
NeurIPS Conference is now Live on Bluesky!
-NeurIPS2024 Communication Chairs
-NeurIPS2024 Communication Chairs
November 22, 2024 at 1:33 AM
NeurIPS Conference is now Live on Bluesky!
-NeurIPS2024 Communication Chairs
-NeurIPS2024 Communication Chairs
If Pranav says it, I believe it
Death of pre-training has been greatly exaggerated.
Look at image generation models. We’re so far from compute optimality yet pre-training is by far the most important bit.
Diffusion models are like o1: train for 1 step but unroll for many. Even test time inference is pre-training bottlenecked.
Look at image generation models. We’re so far from compute optimality yet pre-training is by far the most important bit.
Diffusion models are like o1: train for 1 step but unroll for many. Even test time inference is pre-training bottlenecked.
November 23, 2024 at 6:52 PM
If Pranav says it, I believe it
(1/3) Very interesting development for autonomous driving!
A key part of the case Tesla has been making about their approach (vs Waymo) is that they can bring the cost down by a lot and scale up production/access because they don't use LIDAR.
A key part of the case Tesla has been making about their approach (vs Waymo) is that they can bring the cost down by a lot and scale up production/access because they don't use LIDAR.
Baidu reveals low-cost Level 4 AV for 2023 deployment on Apollo Go
The company says that the Apollo RT6 autonomous vehicle is ready to provide driverless service as the company moves toward a future in which taking a robotaxi will be half the cost of taking a taxi to...
insideautonomousvehicles.com
November 22, 2024 at 9:33 PM
(1/3) Very interesting development for autonomous driving!
A key part of the case Tesla has been making about their approach (vs Waymo) is that they can bring the cost down by a lot and scale up production/access because they don't use LIDAR.
A key part of the case Tesla has been making about their approach (vs Waymo) is that they can bring the cost down by a lot and scale up production/access because they don't use LIDAR.
Amazing PhD opportunity with Jakob (@jfoerst.bsky.social) offering time split between Oxford and FAIR!
Note that the deadline is Dec 2nd!
x.com/j_foerst/sta...
Note that the deadline is Dec 2nd!
x.com/j_foerst/sta...
x.com
x.com
November 22, 2024 at 7:42 PM
Amazing PhD opportunity with Jakob (@jfoerst.bsky.social) offering time split between Oxford and FAIR!
Note that the deadline is Dec 2nd!
x.com/j_foerst/sta...
Note that the deadline is Dec 2nd!
x.com/j_foerst/sta...
Glad this is taking off! I'll be posting a lot more here than the other place (hopefully!)
November 22, 2024 at 6:37 PM
Glad this is taking off! I'll be posting a lot more here than the other place (hopefully!)