Maze action video game Pac-Man
Machine Learning Research

Two-Way Winner

AlphaGo Zero demonstrates superhuman performance playing Go, chess, and shogi. Models like R2D2 do the same playing classic Atari titles. A new approach to deep reinforcement learning is the first to achieve state-of-the-art results playing both board and video games.
Single Headed Attention RNN (SHA-RNN)
Machine Learning Research

Language Modeling on One GPU

The latest large, pretrained language models rely on trendy layers based on transformer networks. New research shows that these newfangled layers may not be necessary.
EfficientDet explained
Machine Learning Research

Easy on the Eyes

Researchers aiming to increase accuracy in object detection generally enlarge the network, but that approach also boosts computational cost. A novel architecture sets a new state of the art in accuracy while cutting the compute cycles required.
Graph related to Noisy Student performance on ImageNet
Machine Learning Research

Self-Training for Sharper Vision

The previous state-of-the-art image classifier was trained on the ImageNet dataset plus 3.5 billion supplemental images from a different database. A new method achieved higher accuracy with one-tenth as many supplemental examples — and they were unlabeled, to boot.
Information related to Implicit Reinforcement without Interaction at Scale (IRIS)
Machine Learning Research

Different Skills From Different Demos

Reinforcement learning trains models by trial and error. In batch reinforcement learning (BRL), models learn by observing many demonstrations by a variety of actors. But what if one doctor is handier with a scalpel while another excels at suturing?
Automatically generated text summary from FactCC with misleading facts highlighted in different colors.
Machine Learning Research

Keeping the Facts Straight: NLP System FactCC Fact Checks Texts

Automatically generated text summaries are becoming common in search engines and news websites. But existing summarizers often mix up facts. For instance, a victim’s name might get switched for the perpetrator’s.
Observational dropout
Machine Learning Research

Seeing the World Blindfolded

In reinforcement learning, if researchers want an agent to have an internal representation of its environment, they’ll build and train a world model that it can refer to. New research shows that world models can emerge from standard training, rather than needing to be built separately.
Information related to Bias-Resilient Neural Network (BR-Net)
Machine Learning Research

Bias Fighter

Sophisticated models trained on biased data can learn discriminatory patterns, which leads to skewed decisions. A new solution aims to prevent neural networks from making decisions based on common biases.
Information about a model for multi-document summarization and question answering
Machine Learning Research

Bigger Corpora, Better Answers

Models that summarize documents and answer questions work pretty well with limited source material, but they can slip into incoherence when they draw from a sizeable corpus. Recent work addresses this problem.
Graph related to LIME and SHAP methods
Machine Learning Research

Bias Goes Undercover

As black-box algorithms like neural networks find their way into high-stakes fields such as transportation, healthcare, and finance, researchers have developed techniques to help explain models’ decisions. New findings show that some of these methods can be fooled.
Comparison between TrXL and GTrXL
Machine Learning Research

Melding Transformers with RL

Large NLP models like BERT can answer questions about a document thanks to the transformer network, a sequence-processing architecture that retains information across much longer sequences than previous methods. But transformers have had little success in reinforcement learning — until now.
Data related to RPDet
Machine Learning Research

Beyond the Bounding Box

Computer vision models typically draw bounding boxes around objects they spot, but those rectangles are a crude approximation of an object’s outline. A new method finds keypoints on an object’s perimeter to produce state-of-the-art object classification.
Information related to a model that predicts a chemical's smell
Machine Learning Research

Nose Job

Predicting a molecule’s aroma is hard because slight changes in structure lead to huge shifts in perception. Good thing deep learning is developing a sense of smell.
Word vectors
Machine Learning Research

Finer Tuning

A word-embedding model typically learns vector representations from a large, general-purpose corpus like Google News. But to make the resulting vectors useful in a specialized domain, they must be fine-tuned on a smaller, domain-specific dataset. Researchers offer a more accurate method.
OctConv example
Machine Learning Research

Convolution Revolution

Looking at images, people see outlines before the details within them. A replacement for the traditional convolutional layer decomposes images based on this distinction between coarse and fine features.

Subscribe to The Batch

Stay updated with weekly AI News and Insights delivered to your inbox