Latest episodes of the podcast Mechanical Dreams
Mostrando página 2 de 5
Scalable-Softmax Is Superior for Attention
11/06/2025
Breast Cancer Recurrence Prediction
06/06/2025
Native Sparse Attention
04/06/2025
Critical Batch Size Revisited
03/06/2025
Rope to Nope and Back Again
16/05/2025
Base of RoPE Bounds Context Length
16/05/2025
SkyLadder
09/05/2025
LLMs on the Line
07/05/2025
The Leaderboard Illusion
30/04/2025
Not All Data Are Unlearned Equally
15/04/2025
A Multi-Power Law for Loss Curve Prediction
14/04/2025
ZARZA We are Zarza, the prestigious firm behind major projects in information technology.