Scholar
JB Lanier
Google Scholar ID: 6IkG2m0AAAAJ
UC Irvine
Citations & Impact
Citations
310
H-index
7
i10-index
6
Publications
18
Co-authors
8
list available
Publications
Resume (English only)
Academic Achievements
- Adapting World Models with Latent-State Dynamics Residuals (Preprint, March 2025)
- Toward Optimal Policy Population Growth in Two-Player Zero-Sum Games (ICLR 2024)
- Selective Perception: Learning Concise State Descriptions for Language Model Actors (NAACL 2024)
- Feasible Adversarial Robust Reinforcement Learning for Underspecified Environments (NeurIPS 2022 Deep RL Workshop)
- Self-Play PSRO: Toward Optimal Populations in Two-Player Zero-Sum Games (arXiv, 2022)
- Anytime PSRO for Two-Player Zero-Sum Games (arXiv, 2022)
- XDO: A Double Oracle Algorithm for Extensive-Form Games (NeurIPS 2021)
- Improving Social Welfare While Preserving Autonomy via a Pareto Mediator (arXiv, 2021)
- OffWorld Gym: Open-Access Physical Lunar Analog Environment for Reinforcement Learning and Robotics Research (COSPAR 2021)
- Pipeline PSRO: A Scalable Approach for Finding Approximate Nash Equilibria in Large Games (NeurIPS 2020, equal contribution)
- ColosseumRL: A Framework for Multiagent Reinforcement Learning in N-Player Games (COMARL AAAI 2020)
- Curiosity-Driven Multi-Criteria Hindsight Experience Replay (NeurIPS 2019 Deep RL Workshop)
Background
- PhD Student at UC Irvine
- Specializes in Deep Reinforcement Learning for multi-agent systems and robotics
- Recent work focuses on efficiently training strong agents for two-player competitive games
- Currently researching simulation-to-real transfer for robotic control
- Developing new model-based RL methods that generate plannable world models for novel environments
- Aims to build reliable agents capable of reacting to unexpected situations
