Forager is a lightweight partially-observable continual RL environment that exposes loss of plasticity in current agents and highlights the value of state construction for ongoing learning.
hub Canonical reference
Title resolution pending
Canonical reference. 86% of citing Pith papers cite this work as background.
19
Pith papers citing it
Background
86% of classified citations
abstract
DeepMind Lab is a first-person 3D game platform designed for research and development of general artificial intelligence and machine learning systems. DeepMind Lab can be used to study how autonomous artificial agents may learn complex tasks in large, partially observed, and visually diverse worlds. DeepMind Lab has a simple and flexible API enabling creative task-designs and novel AI-designs to be explored and quickly iterated upon. It is powered by a fast and widely recognised game engine, and tailored for effective use by the research community.
hub tools
citation-role summary
background 5
dataset 2
citation-polarity summary
representative citing papers
DreamerV3 uses world models and robustness techniques to solve over 150 tasks across domains with a single configuration, including Minecraft diamond collection from scratch.
Gato is a multi-modal, multi-task, multi-embodiment generalist policy using one transformer network to handle text, vision, games, and robotics tasks.
Dreamer learns to control from images by imagining and optimizing behaviors in a learned latent world model, outperforming prior methods on 20 visual tasks in data efficiency and final performance.
The paper surveys CRL literature, proposes a taxonomy of methods into four categories based on knowledge storage and transfer, reviews metrics and benchmarks, and outlines challenges and future research directions.
Compressive Transformer sets new records on WikiText-103 (17.1 ppl) and Enwik8 (0.97 bpc) via memory compression and introduces the PG-19 long-range language benchmark.
VTAB is a 19-task benchmark that measures representation quality by few-shot adaptation performance across diverse vision domains, with a controlled large-scale comparison of popular pretraining methods.
Arena introduces a modular Interface design that extends OpenAI Gym wrappers to support complex multi-agent RL scenarios including self-play and cooperative-competitive interactions.
MDPD mutually distills knowledge between a frozen backbone and a learnable side network during fine-tuning, then discards the side network at inference to accelerate speed by at least 25% while preserving accuracy.
GameWorld is a new benchmark providing standardized interfaces, 34 games, 170 tasks, and verifiable outcome metrics to evaluate multimodal large language model agents in video game environments.
Activation prompts on intermediate layers outperform input-level visual prompting and parameter-efficient fine-tuning in accuracy and efficiency across 29 datasets.
LLaVA-Video-178K is a new synthetic video instruction dataset that, when combined with existing data to train LLaVA-Video, produces strong results on video understanding benchmarks.
Consensus recommendations for standardized evaluation measures, problem statements, and benchmarking scenarios in embodied navigation research.
An empirical literature analysis reveals a bifurcation in RL environments into Semantic Prior (LLM-dominated) and Domain-Specific Generalization ecosystems with distinct cognitive fingerprints.
PVeRA extends VeRA by making its frozen random low-rank matrices probabilistic, enabling better handling of ambiguities and outperforming prior adapters on the VTAB-1k benchmark.
Multi-step predictive generative models form stable belief states capturing environment layout and agent pose, yielding higher data efficiency on RL tasks than model-free agents.
A rationale is presented for developing an assistant in Minecraft to advance natural language understanding and dialogue learning.
Adaptive replacements for domain-specific components in deep RL agents can yield better learning on new tasks without additional tuning.
Experiments indicate deep learning models achieve higher accuracy on numerosity tasks for counts below human subitizing capacity.
citing papers explorer
-
Forager: a lightweight testbed for continual learning with partial observability in RL
Forager is a lightweight partially-observable continual RL environment that exposes loss of plasticity in current agents and highlights the value of state construction for ongoing learning.
-
Mastering Diverse Domains through World Models
DreamerV3 uses world models and robustness techniques to solve over 150 tasks across domains with a single configuration, including Minecraft diamond collection from scratch.
-
A Generalist Agent
Gato is a multi-modal, multi-task, multi-embodiment generalist policy using one transformer network to handle text, vision, games, and robotics tasks.
-
Dream to Control: Learning Behaviors by Latent Imagination
Dreamer learns to control from images by imagining and optimizing behaviors in a learned latent world model, outperforming prior methods on 20 visual tasks in data efficiency and final performance.
-
A Survey of Continual Reinforcement Learning
The paper surveys CRL literature, proposes a taxonomy of methods into four categories based on knowledge storage and transfer, reviews metrics and benchmarks, and outlines challenges and future research directions.
-
Compressive Transformers for Long-Range Sequence Modelling
Compressive Transformer sets new records on WikiText-103 (17.1 ppl) and Enwik8 (0.97 bpc) via memory compression and introduces the PG-19 long-range language benchmark.
-
A Large-scale Study of Representation Learning with the Visual Task Adaptation Benchmark
VTAB is a 19-task benchmark that measures representation quality by few-shot adaptation performance across diverse vision domains, with a controlled large-scale comparison of popular pretraining methods.
-
Arena: a toolkit for Multi-Agent Reinforcement Learning
Arena introduces a modular Interface design that extends OpenAI Gym wrappers to support complex multi-agent RL scenarios including self-play and cooperative-competitive interactions.
-
Memory-Efficient Transfer Learning with Fading Side Networks via Masked Dual Path Distillation
MDPD mutually distills knowledge between a frozen backbone and a learnable side network during fine-tuning, then discards the side network at inference to accelerate speed by at least 25% while preserving accuracy.
-
GameWorld: Towards Standardized and Verifiable Evaluation of Multimodal Game Agents
GameWorld is a new benchmark providing standardized interfaces, 34 games, 170 tasks, and verifiable outcome metrics to evaluate multimodal large language model agents in video game environments.
-
Visual prompting reimagined: The power of the Activation Prompts
Activation prompts on intermediate layers outperform input-level visual prompting and parameter-efficient fine-tuning in accuracy and efficiency across 29 datasets.
-
LLaVA-Video: Video Instruction Tuning With Synthetic Data
LLaVA-Video-178K is a new synthetic video instruction dataset that, when combined with existing data to train LLaVA-Video, produces strong results on video understanding benchmarks.
-
On Evaluation of Embodied Navigation Agents
Consensus recommendations for standardized evaluation measures, problem statements, and benchmarking scenarios in embodied navigation research.
-
From Pixels to Digital Agents: An Empirical Study on the Taxonomy and Technological Trends of Reinforcement Learning Environments
An empirical literature analysis reveals a bifurcation in RL environments into Semantic Prior (LLM-dominated) and Domain-Specific Generalization ecosystems with distinct cognitive fingerprints.
-
PVeRA: Probabilistic Vector-Based Random Matrix Adaptation
PVeRA extends VeRA by making its frozen random low-rank matrices probabilistic, enabling better handling of ambiguities and outperforming prior adapters on the VTAB-1k benchmark.
-
Shaping Belief States with Generative Environment Models for RL
Multi-step predictive generative models form stable belief states capturing environment layout and agent pose, yielding higher data efficiency on RL tasks than model-free agents.
-
Why Build an Assistant in Minecraft?
A rationale is presented for developing an assistant in Minecraft to advance natural language understanding and dialogue learning.
-
On Inductive Biases in Deep Reinforcement Learning
Adaptive replacements for domain-specific components in deep RL agents can yield better learning on new tasks without additional tuning.
-
On the notion of number in humans and machines
Experiments indicate deep learning models achieve higher accuracy on numerosity tasks for counts below human subitizing capacity.