ΑΙhub.org

deep dive


Tackling diverse tasks with neural architecture search

We developed a Neural Architecture Search method that generates and trains task-specific convolutional neural networks.
24 October 2022, by

Tracking any pixel in a video

We propose Persistent Independent Particles (PIPs), a new particle video method to track pixels in a video.
17 October 2022, by

Keeping learning-based control safe by regulating distributional shift

We propose a new framework to reason about the safety of a learning-based controller with respect to its training distribution.
30 September 2022, by

Recurrent model-free RL can be a strong baseline for many POMDPs

Considering an approach for dealing with realistic problems with noise and incomplete information.
23 September 2022, by

Reverse engineering the NTK: towards first-principles architecture design

We propose a paradigm for bringing some principle to the art of architecture design.
12 September 2022, by

Galaxies on graph neural networks

Using Graph Neural Networks, we trained Generative Adversarial Networks to correctly predict the coherent orientations of galaxies in a state-of-the-art cosmological simulation.
05 September 2022, by


AIhub is supported by:







auton-survival: An open-source package for regression, counterfactual estimation, evaluation and phenotyping censored time-to-event data

We present auton-survival – a comprehensive Python code repository of user-friendly, machine learning tools for working with censored time-to-event data.
22 August 2022, by

Why do policy gradient methods work so well in cooperative MARL? Evidence from policy representation

We show how policy gradient methods can converge to an optimal policy in certain cases, and that they can learn multi-modal policies.
15 August 2022, by

Does AutoML work for diverse tasks?

Can the available AutoML tools quickly and painlessly attain near-expert performance on diverse learning tasks?
01 August 2022, by

FIGS: Attaining XGBoost-level performance with the interpretability and speed of CART

In this blog post we cover a new method for fitting an interpretable model that takes the form of a sum of trees.
12 July 2022, by

Deep attentive variational inference

The expressivity of current deep probabilistic models can be improved by selectively prioritizing statistical dependencies between latent variables that are potentially distant from each other.
24 June 2022, by

Rethinking human-in-the-loop for artificial augmented intelligence

How do we build and evaluate an AI system for real-world applications?
17 June 2022, by

Bootstrapped meta-learning – an interview with Sebastian Flennerhag

ICLR2022 award winner tells us about how he and co-authors approached the meta-learning problem.
07 June 2022, by

Designing societally beneficial reinforcement learning systems

Studying the risks associated with using reinforcement learning for real-world applications.
31 May 2022, by

An experimental design perspective on model-based reinforcement learning

We propose a simple algorithm that is able to solve a wide variety of control tasks.
19 May 2022, by

Should I use offline RL or imitation learning?

In this blog post, we aim to understand if, when and why offline RL is a better approach for tackling a variety of sequential decision-making problems.
17 May 2022, by

Offline RL made easier: no TD learning, advantage reweighting, or transformers

We try to identify the essential elements of offline RL via supervised learning.
03 May 2022, by

Unsupervised skill discovery with contrastive intrinsic control

Unsupervised reinforcement learning (RL), where RL agents pre-train with self-supervised rewards, is an emerging paradigm for developing RL agents that are capable of generalization.
01 April 2022, by

Assessing generalization of SGD via disagreement

We demonstrate that a simple procedure can accurately estimate the generalization error with only unlabeled data.
21 March 2022, by

imodels: leveraging the unreasonable effectiveness of rules

imodels provides a simple unified interface and implementation for many state-of-the-art interpretable modeling techniques, particularly rule-based methods.
14 March 2022, by

Why spectral normalization stabilizes GANs: analysis and improvements

We investigate the training stability of generative adversarial networks (GANs).
07 March 2022, by

The unsupervised reinforcement learning benchmark

We consider the unsupervised RL problem - how do we learn useful behaviors without supervision and then adapt them to solve downstream tasks quickly?
14 February 2022, by

Improving RL with lookahead: learning off-policy with online planning

We suggest using a policy that looks ahead using a learned model to find the best action sequence.
11 February 2022, by

Sequence modeling solutions for reinforcement learning problems

We tackle large-scale reinforcement learning problems with the toolbox of sequence modeling.
03 February 2022, by

An energy-based perspective on learning observation models

We propose a conceptually novel approach to mapping sensor readings into states.
21 January 2022, by

Which mutual information representation learning objectives are sufficient for control?

How can we best design representation learning objectives?
10 January 2022, by

Understanding user interfaces with screen parsing

We introduce the problem of screen parsing, which we use to predict structured user interface models from visual information.
03 January 2022, by

Bridge data: boosting generalization of robotic skills with cross-domain datasets

With our proposed dataset and multi-task, multi-domain learning approach, we have shown one potential avenue for making diverse datasets reusable in robotics.
30 December 2021, by

Why generalization in RL is difficult: epistemic POMDPs and implicit partial observability

In this blog post, we will aim to explain why generalization in RL is fundamentally hard, even in theory.
21 December 2021, by

Designs from data: offline black-box optimization via conservative training

In this post, we discuss offline model-based optimization and some recent advances in this area.
10 December 2021, by







©2021 - Association for the Understanding of Artificial Intelligence


 












©2021 - ROBOTS Association