(From Lei et al, “Geometric Understanding of Deep Learning,” (2018), 1805.10451)
Posts
An introduction to DSPy, the PyTorch for LLM programs
Feb 10, 2024
Chunking strategies for increasing effectiveness of RAG systems
Aug 25, 2023
Large Language Model (LLM) fine-tuning with low-rank matrix approximations and data-type quantization.
Jun 22, 2023
Efficient and Effective Passage Search via
Contextualized Late Interaction over BERT
Mar 7, 2022
Understanding the new type of position encoding that unifies absolute and relative approaches.
Feb 19, 2022
Trying to make sense of interpolation, generalization, and the theoretical foundations that connect them like the Universal Approximation Theory.
Dec 11, 2021
When neural networks and more general non-linear models are accurately approximated by their linearizations
Oct 2, 2021
Understanding one of the most effective and useful process for generative modeling.
Feb 1, 2020
A rigorous proof of the convergence rate of PGM, and code implementation from scratch using MNIST data
Dec 5, 2019
From scratch implementation of accelerated GD and Newton's method using a funky gamma-distributed loss function
Nov 28, 2019
Understanding and applying the novel Focal Loss for classification with extreme class imbalances.
Mar 11, 2019
