MACHINE LEARNING

Table-LLaVa 7B: Multimodal Table Understanding Competing with GPT-4V

By

–

16 June 2024 19h16

9/ Multimodal Table Understanding – introduces Table-LLaVa 7B, a multimodal LLM for multimodal table understanding; it’s competitive with GPT-4V and significantly outperforms existing MLLMs on multiple benchmarks.

Mixture of Memory Experts Reduces LLM Hallucination Tenfold

By

–

16 June 2024 19h16

8/ Mixture of Memory Experts – proposes an approach to significantly reduce hallucination (10x) by tuning millions of expert adapters (e.g., LoRAs) to learn exact facts and retrieve them from an index at inference time.

Transformers Meet Neural Algorithmic Reasoners Hybrid Architecture

By

–

16 June 2024 19h16

5/ Transformers Meet Neural Algorithmic Reasoners – a new hybrid architecture that enables tokens in the LLM to cross-attend to node embeddings from a GNN-based neural algorithmic reasoner (NAR).

LLM-Driven Discovery of Preference Optimization Algorithms

By

–

16 June 2024 19h16

2/ Discovering Preference Optimization Algorithms with LLMs – proposes LLM-driven objective discovery of state-of-the-art preference optimization.https://t.co/tXYiVCzW9S
— DAIR.AI (@dair_ai) 16 juin 2024

2/ Discovering Preference Optimization Algorithms with LLMs – proposes LLM-driven objective discovery of state-of-the-art preference optimization.

→ View original post on X — @jeande_d

Deep Learning Computer Vision Repository Goals and Original Vision

By

@jeande_d

–

16 June 2024 18h19

The goal of working on Deep Learning for Computer Vision was to have a practical repo for people in deep learning and visual recognition, serving the role Complete Machine Learning Package does in ML foundations. It's not what I originally envisioned(will explain), but the

→ View original post on X — @jeande_d

Deep Learning for Computer Vision V1 Release Announced

By

@jeande_d

–

16 June 2024 18h19

Releasing Deep Learning for Computer Vision V1 Inspired by Complete Machine Learning Package, Deep Learning for Computer Vision is a comprehensive repository/knowledge base of topics in deep learning and computer vision. The repository covers various topics including but

→ View original post on X — @gdb

Following Curiosity in Machine Learning Engineering Debugging

By

@gdb

–

16 June 2024 3h00

a hard but very fun part of machine learning engineering is following your own curiosity to chase down every unexplained detail of system behavior