Creează un cont pentru a te abona la acest blog și a-i urmări articolele pe pagina Lectură.
-
I swapped the transformer's matchmaking service for a handful of Gaussians, and it beat content scoring without comparing a single pair of tokens.
-
The standard Cross-Entropy loss has a well-known flaw: it forces neural networks to push their features toward infinity just to reach full confidence.
-
If you crack open the architecture of almost any modern Transformer, you will find Scaled Dot-Product Attention (SDPA) sitting at its core.
-
The "Infinite Gap" and Why Softmax Keeps Me Up at Night To understand any solution, we first have to really understand the problem.
-
Classification with neural networks is weird!
-
A lot has happened since my last post on the Sharpened Cosine Similarity layer.
-
Some days ago Brandon Rohrer retweeted his own twitter thread from 2020 in which he makes the argument that convolutions are actually pretty bad at extracting features.
-
CLIP is a model published by OpenAI that is able to learn visual concepts by natural language supervision.
-
Image augmentations make all the difference when working with neural networks.
-
Since my paper on "Joint Unsupervised Depth-Estimation and Obstacle-Detection" did not get accepted to NeuRIPS 2019 I now had another unpublished paper lying around.