Hi [First name],
We are hiring a designer to join us full-time in SF! Help take alphaXiv’s design to the next level and make research easier to explore for millions across the globe.
Upcoming events
|
Chris Manning x David Ha x Ekin Dogus Panel @ NeurIPS
Tuesday, December 8th 10:00 PM PT
Join alphaXiv and Canva at NeurIPS on December 9 for a panel on recursive self-improvement and scientific discovery with Chris Manning (Stanford AI Lab Director), David Ha (Sakana AI CEO), and Ekin Dogus Cubuk (Periodic Labs Cofounder). Stay afterward for drinks and conversation.
Sign Up Here
|
|
Trending papers
|
View all
|
Unbundling Multimodal Vision Encoding
How multimodal language models should encode visual input remains up for debate. Most models use a pretrained vision encoder, while some recent recipe...
2,945 views
| |
An Energy-Based Perspective on Machine Learning
Energy-based models (EBMs) define a distribution through a scalar energy, $p_\theta(x)\propto\exp(-E_\theta(x))$, and are usually presented as one mod...
2,781 views
|
|
Latent Feedback Is a Compute and Data Amplifier: Scaling Law for Latent Feedback Transf...
When high-quality data runs out, a common practice is to train on it for several epochs, but each repeat is worth less than the last. Latent feedback,...
2,276 views
| |
SLAM Dunk: A Recipe for Long-Horizon Egocentric SLAM in the Wild
Standard visual-inertial odometry (VIO) systems struggle to maintain accuracy with long trajectories in the wild, especially in egocentric settings, a...
2,264 views
|
|
Interpretable Weight-Space Programming Builds Transparent Transformers and Brain Encodi...
Neural networks are powerful predictors, yet their opacity limits their usefulness in science, where the goal is explanation rather than mere predicti...
5,076 views
| |
Weight-Space Reading
Can a language model predict what an edit to a network’s weights will do to its behavior, without running the network? We call this weight-space readi...
2,018 views
|
|
|