Vitor Meriat
banner
vitormeriat.bsky.social
Vitor Meriat
@vitormeriat.bsky.social
ML Engineer | Python 𓆙 | Research 🔬
Aurélio Buarque de Holanda Ferreira, dicionarista e lexicógrafo responsável pelo famoso dicionário Aurélio.
Aurélio Buarque de Holanda: como se faz um dicionário, 1976
YouTube video by Memória Audiovisual Brasileira - MAB
www.youtube.com
March 13, 2025 at 9:59 AM
(1º) EUA faz embargo de GPUs há anos (2º) Tentativa de limitar a concorrência (3º) OpenAi jogando ao vento que a China fez uso de GPUs modernas para alcançar o resultado do #DeepSeek. Isso É querer limitar avanços na ciência baseado em sua incopetência. www.bloomberg.com/news/article...
January 29, 2025 at 10:16 AM
Reposted by Vitor Meriat
Graph Transformers (GTs) can handle long-range dependencies and resolve information bottlenecks, but they’re computationally expensive. Our new model, Spexphormer, helps scale them to much larger graphs – check it out at NeurIPS next week, or the preview here!
[1/13]
#NeurIPS2024
December 5, 2024 at 7:58 PM
Reposted by Vitor Meriat
What's the difference between explainability and interpretability? Does the machine learning community have an agreed-upon definition?

No.

There's a great overview, which is from this paper: arxiv.org/abs/2211.08943

My take: I prefer interpretability since the term explainability is too strong.
November 26, 2024 at 3:41 PM
Reposted by Vitor Meriat
TL;DR gradient descent!

From fleuret.org/dlc/material...
November 26, 2024 at 6:31 AM
The library consists of various dynamic and temporal geometric deep learning, embedding, and spatio-temporal regression methods from a variety of published research papers. #GNN #GraphNeuralNetwork
GitHub - benedekrozemberczki/pytorch_geometric_temporal: PyTorch Geometric Temporal: Spatiotemporal Signal Processing with Neural Machine Learning Models (CIKM 2021)
PyTorch Geometric Temporal: Spatiotemporal Signal Processing with Neural Machine Learning Models (CIKM 2021) - benedekrozemberczki/pytorch_geometric_temporal
github.com
November 25, 2024 at 11:31 PM