tdooms
Papers
A magnifying glass

Bilinear Decomposition

(Jun 2024)

Using bilinear MLPs to reverse-engineer shallow MNIST and Tiny Stories models using only their weights.

A split arrow

Tokenized SAEs

(Jun 2024)

We propose using a per-token bias in SAEs to separate token reconstructions from interesting, semantic features.

A trident

The Trifecta

(Nov 2023)

Three techniques to significantly improve the Forward-Forward algorithm. We achieve 84% on CIFAR-10.

And much more to come...

Presentations

Weight-based interpretability

(Oct 2024)

A brief introduction to weight-based interpretability.

Introduction to Transformers

(Jan 2024)

An informal introduction to the Transformer architecture.

Embodied Language Models

(Jun 2023)

A presentation on the integration of LLMs into robotics.

Diffusion Models

(Jan 2023)

A brief overview of diffusion models and the current SOTA.