VLRS-Bench is the first benchmark dedicated to complex vision-language reasoning in remote sensing, with 2000 QA pairs across 14 tasks in cognition, decision, and prediction dimensions.
hub Canonical reference
Dota 2 with Large Scale Deep Reinforcement Learning
Canonical reference. 93% of citing Pith papers cite this work as background.
78
Pith papers citing it
Background
93% of classified citations
abstract
On April 13th, 2019, OpenAI Five became the first AI system to defeat the world champions at an esports game. The game of Dota 2 presents novel challenges for AI systems such as long time horizons, imperfect information, and complex, continuous state-action spaces, all challenges which will become increasingly central to more capable AI systems. OpenAI Five leveraged existing reinforcement learning techniques, scaled to learn from batches of approximately 2 million frames every 2 seconds. We developed a distributed training system and tools for continual training which allowed us to train OpenAI Five for 10 months. By defeating the Dota 2 world champion (Team OG), OpenAI Five demonstrates that self-play reinforcement learning can achieve superhuman performance on a difficult task.
hub tools
citation-role summary
background 13
other 1
citation-polarity summary
representative citing papers
Generative agents with memory streams, reflection, and planning using LLMs exhibit believable individual and emergent social behaviors in a simulated town.
GPT-f, a transformer-based prover for Metamath, generated new short proofs that were accepted into the main library—the first such contribution from a deep-learning system.
GPTNT benchmark demonstrates that state-of-the-art multimodal models cannot perform real-time collaborative bomb defusal in Keep Talking and Nobody Explodes, unlike human players.
Concept-based models can use controlled 'benign' information leakage to remain accurate and intervenable under real-world concept incompleteness by reframing their training objective.
RAT reformulates regularized natural policy gradients as vanilla gradients with a transformed advantage, computed efficiently via randomized block Kaczmarz iterations on on-policy data.
A projected gradient descent algorithm for noisy inductive matrix completion achieves linear convergence and stable recovery at sample complexity governed by side-information dimension, extending to inexact side-information with optimal error degradation.
DRATS derives a minimax objective from a feasibility formulation of MTRL to adaptively sample tasks with the largest return gaps, leading to better worst-task performance on MetaWorld benchmarks.
Standard MORL metrics do not measure whether preference inputs reliably control agent behavior, so a new controllability metric is introduced to restore the link between user intent and agent output.
LC-MAPF uses multi-round local communication between neighboring agents in a pre-trained model to outperform prior learning-based MAPF solvers on diverse unseen scenarios while preserving scalability.
Synthetic data augmentation helps channel-mixing time series models but degrades channel-independent ones, with reliable gains only from seasonal-trend generators and gradual schedules in low-resource settings.
COSPLAY co-evolves an LLM decision agent with a skill bank agent to improve long-horizon game performance, reporting over 25.1% average reward gains versus frontier LLM baselines on single-player benchmarks.
InfoChess proposes a symmetric adversarial game focused purely on information control and probabilistic king-location inference, with RL agents outperforming heuristic baselines and gameplay dissected via belief entropy, cross-entropy, and predictive scores.
PPO in a new competitive game fails due to five implementation bugs and then competitive overfitting where self-play stays near 50% but generalization drops to 21.6%; mixing 20% random opponents restores generalization to 77.1%.
NePPO learns a player-independent potential function via a novel objective whose minimization yields an approximate Nash equilibrium for general-sum multi-agent games.
Information geometry constrains intrinsic rewards to strictly concave functions of reciprocal occupancy, with geodesic interpolation on the occupancy manifold yielding a scalar-parameter family that includes count-based and max-entropy exploration.
Voyager achieves superior lifelong learning in Minecraft by combining an automatic exploration curriculum, a library of executable skills, and iterative LLM prompting with environment feedback, yielding 3.3x more unique items and 15.3x faster milestone unlocks than prior methods while generalizing技能
Introduces the Dota2-Vis dataset of 288 videos from 144 TI 2025 matches plus 2,477 annotated minimaps and evaluates YOLO11 variants for player-icon detection to produce visibility curves.
EMAgnet replaces uniform-magnet regularization in PPO self-play with an EMA of last-iterate policy parameters and reports lower exploitability on most tested zero-sum benchmarks, especially those with dominated strategies.
Self-play RL with a vision transformer policy, powered by a 10,000x faster JAX simulator, produces an agent that ranks #1 on the Generals.io leaderboard and wins 199-70 against top humans.
Asymmetric physics (high-fidelity non-diff simulator plus differentiable surrogates) enables end-to-end training of decentralized vision-based policies for up to 512 quadrupeds that transfer zero-shot to real hardware.
TSP reframes secure code generation as a tree-structured self-play process that supplies dense on-policy signals at vulnerability-prone nodes, yielding higher security pass rates and cross-language generalization than SFT or unstructured self-play.
Adversarial co-evolution of LLM constitutions in public goods games reaches near-parity equilibrium only when fitness is coupled across factions and evaluation uses at least five seeds per generation.
pcsp is a shared RL policy using LLM persona embeddings, low-rank projection, and PPO+InfoNCE+KL training that delivers 17x above-chance zero-shot persona identification and 22x faster inference on a 300-persona benchmark.
citing papers explorer
-
In Defense of Information Leakage in Concept-based Models
Concept-based models can use controlled 'benign' information leakage to remain accurate and intervenable under real-world concept incompleteness by reframing their training objective.
-
Randomized Advantage Transformation (RAT): Computing Natural Policy Gradients via Direct Backpropagation
RAT reformulates regularized natural policy gradients as vanilla gradients with a transformed advantage, computed efficiently via randomized block Kaczmarz iterations on on-policy data.
-
Distributionally Robust Multi-Task Reinforcement Learning via Adaptive Task Sampling
DRATS derives a minimax objective from a feasibility formulation of MTRL to adaptively sample tasks with the largest return gaps, leading to better worst-task performance on MetaWorld benchmarks.
-
Controllability in preference-conditioned multi-objective reinforcement learning
Standard MORL metrics do not measure whether preference inputs reliably control agent behavior, so a new controllability metric is introduced to restore the link between user intent and agent output.
-
Does Synthetic Data Help? Empirical Evidence from Deep Learning Time Series Forecasters
Synthetic data augmentation helps channel-mixing time series models but degrades channel-independent ones, with reliable gains only from seasonal-trend generators and gradual schedules in low-resource settings.
-
Territory Paint Wars: Diagnosing and Mitigating Failure Modes in Competitive Multi-Agent PPO
PPO in a new competitive game fails due to five implementation bugs and then competitive overfitting where self-play stays near 50% but generalization drops to 21.6%; mixing 20% random opponents restores generalization to 77.1%.
-
NePPO: Near-Potential Policy Optimization for General-Sum Multi-Agent Reinforcement Learning
NePPO learns a player-independent potential function via a novel objective whose minimization yields an approximate Nash equilibrium for general-sum multi-agent games.
-
EMAgnet: Parameter-Space EMA Regularization for Policy Gradient Self-Play in Large Games
EMAgnet replaces uniform-magnet regularization in PPO self-play with an EMA of last-iterate policy parameters and reports lower exploitability on most tested zero-sum benchmarks, especially those with dominated strategies.
-
Superhuman AI for Generals.io Using Self-Play Reinforcement Learning
Self-play RL with a vision transformer policy, powered by a 10,000x faster JAX simulator, produces an agent that ranks #1 on the Generals.io leaderboard and wins 199-70 against top humans.
-
GAE Falls Short in Imperfect-Information Self-Play Reinforcement Learning
GAE suffers from amplified variance in imperfect-info self-play RL; VRPO with Q-boosting and multi-step Expected SARSA(λ) reduces it and improves performance on mid-to-large games.
-
SOPE: Stabilizing Off-Policy Evaluation for Online RL with Prior Data
SOPE dynamically controls offline training length in online RL using actor-aligned OPE on validation data to stop when benefits saturate, achieving up to 45.6% better performance and 22x less computation on Minari tasks.
-
QHyer: Q-conditioned Hybrid Attention-mamba Transformer for Offline Goal-conditioned RL
QHyer replaces return-to-go with a state-conditioned Q-estimator and adds a gated hybrid attention-mamba backbone to achieve state-of-the-art performance in offline goal-conditioned RL on both Markovian and non-Markovian datasets.
-
Odysseus: Scaling VLMs to 100+ Turn Decision-Making in Games via Reinforcement Learning
Odysseus adapts PPO with a turn-level critic and leverages pretrained VLM action priors to train agents achieving at least 3x average game progress over frontier models in long-horizon Super Mario Land.
-
When Errors Can Be Beneficial: A Categorization of Imperfect Rewards for Policy Gradient
Certain errors in proxy rewards for policy gradient methods can be benign or beneficial by preventing policies from stalling on outputs with mediocre ground truth rewards, enabling improved RLHF metrics and reward design insights.
-
Biased Dreams: Limitations to Epistemic Uncertainty Quantification in Latent Space Models
Latent transitions in models like Dreamer are biased toward dense regions, creating attractors that hide true dynamics discrepancies and cause epistemic uncertainty to be unreliable while overestimating rewards.
-
Learning to Distributedly Estimate under Partially Known Dynamics: A Covariance-Agnostic Neural Kalman Consensus Filter
CA-NKCF is a hybrid neural-Kalman consensus filter for distributed state estimation that operates without noise covariance knowledge and shows robustness to model misspecification in linear, chaotic, and wireless scenarios.
-
Uncertainty-aware reinforcement learning for chemical language models
Uncertainty-aware RL for chemical language models raises true hit rate from 0.5 to 0.75 by favoring low-uncertainty regions during optimization.
-
Direct Advantage Estimation for Scalable and Sample-efficient Deep Reinforcement Learning
Extends DAE theory to POMDPs with minimal changes and introduces discrete latent dynamics to cut computational cost, with ALE experiments showing scalability and retained sample efficiency.
-
An Agency-Transferring Model-Free Policy Enhancement Technique
A model-free RL method arbitrates between a functional baseline policy and a learning policy, transferring agency over time to yield a standalone policy with high goal-reaching rates and competitive returns on continuous-control tasks.
-
Local Guidance, Global Impact: Gaussian-Reshaped Trust Region Unlocks Behavior Transitions
GTR introduces a bounded non-monotonic Gaussian trust region and Mixture Gaussian Anchor to enable effective behavior transitions in non-stationary RL where standard PPO fails.
-
Data-Augmented Game Starts for Accelerating Self-Play Exploration in Imperfect Information Games
DAGS initializes policy-gradient self-play from human-derived intermediate states to reduce exploitability in challenging imperfect-information games, with a multi-task flag fix for resulting bias and new benchmark environments.
-
Position: Deployed Reinforcement Learning should be Continual
Deployed RL agents receiving evaluative rewards face inherent non-stationarity and should engage in continual learning rather than following a train-then-fix approach.
- ARROW: Augmented Replay for RObust World models