Learning From Mistakes Makes LLM Better Reasoner

New techniques efficiently accelerate sparse tensors for massive AI models

🥇Top ML Papers of the Week

Machine Learning Provides a Clearer Window into Ocean Motion - Eos

🥇Top ML Papers of the Week

🥇Top ML Papers of the Week

The Alignment Problem: Machine Learning and Human Values with Brian Christian

What's Botsonic?

🥇Top ML Papers of the Week

François Chollet on X

100x Efficiency: MIT’s Machine-Learning System Based on Light Could Yield More Powerful Large Language Models

🥇Top ML Papers of the Week

Large Language Model as Attributed Training Data Generator: A Tale of Diversity and Bias

Towards Measuring the Representation of Subjective Global Opinions in Language Models

StableRep: Synthetic Images from Text-to-Image Models Make Strong Visual Representation Learners

Quantum neural network cost function concentration dependency on the parametrization expressivity

LLM-Blender: Ensembling Large Language Models with Pairwise Ranking and Generative Fusion

🥇Top ML Papers of the Week

Gong, Y., Rouditchenko, A., Liu, A. H., Harwath, D., Karlinsky, L., Kuehne, H., & Glass, J. (2022). Contrastive audio-visual masked autoencoder. arXiv preprint arXiv:2210.07839.

LEACE: Perfect linear concept erasure in closed form

Enabling Scalable AI Computational Lithography with Physics-Inspired Models | Research

Faith and Fate: Limits of Transformers on Compositionality

CodeTF: One-stop Transformer Library for State-of-the-art Code LLM

Bytes Are All You Need: Transformers Operating Directly On File Bytes

Fine-Tuning Language Models with Just Forward Passes

The Impact of Positional Encoding on Length Generalization in Transformers

MERT: Acoustic Music Understanding Model with Large-Scale Self-supervised Training

Thought Cloning: Learning to Think while Acting by Imitating Human Thinking

BiomedGPT: A Unified and Generalist Biomedical Generative Pre-trained Transformer for Vision, Language, and Multimodal Tasks

Neuralangelo: High-Fidelity Neural Surface Reconstruction | Research