arXiv subjects

Joonyoung Lim

Publications and source records attributed to Joonyoung Lim.

3 recordsLinked to original sources

arXiv2026

Accelerating Safe Reinforcement Learning with Massive Parallelism

We investigate the problem of scaling safe reinforcement learning (safe RL) to massively parallel training regimes, where a large number of environments are exe- cuted with short rollout horizons. Although this setting improves data throughput and wall-clock efficiency, it creates a mismatch with constrained Markov decision processes (CMDPs), whose safety constraints are defined over full-episode cost returns. We address this mismatch by analyzing staggered environment resets as a phase-mixture distribution over short trajectory segments. This analysis charac- terizes the bias induced by phase coverage and motivates a phase-aggregated cost estimator that reconstructs episode-level cost estimates from short rollout batches. We further incorporate finite-sample uncertainty and distribution mismatch into a constraint-tightening framework, yielding a scalable Safe RL method compatible with on-policy optimization. Using an MJX-based implementation of Safety Gym, our experiments show that the proposed method achieves performance comparable to non-massively parallel implementations while substantially reducing wall-clock training time.

cs.LG↗

arXiv2026

LOCU: Lowdin-Orthogonalized Constraint Updates for Multi-Constraint Policy Optimization

We investigate multi-constraint policy optimization in constrained Markov decision processes (CMDPs), where interactions among constraint gradients often yield ill-conditioned update directions, causing numerically fragile steps and unintended cancellation or redundant overlap in constraint corrections. To address this, we propose LOCU, which mitigates numerical instability arising from constraint interactions by applying symmetric Lowdin orthogonalization to the natural gradients of the reward and constraints under the Fisher-Rao metric. Combined with Fisher-geometric screening and near-collinear compression, LOCU yields a compact reduced space in which the trust region becomes a Euclidean ball and constraints are handled symmetrically through a low-dimensional active-set solve. For infeasible iterates, LOCU constructs a Pareto-descent direction which preserves the local feasibility of satisfied constraints within the same reduced formulation. Experiments show that LOCU remains stable under high constraint coupling, narrow feasible regions, and infeasible starts, and generalizes across different cost critic architectures.

cs.AI↗

arXiv2025

KFCPO: Kronecker-Factored Approximated Constrained Policy Optimization

We propose KFCPO, a novel Safe Reinforcement Learning (Safe RL) algorithm that combines scalable Kronecker-Factored Approximate Curvature (K-FAC) based second-order policy optimization with safety-aware gradient manipulation. KFCPO leverages K-FAC to perform efficient and stable natural gradient updates by approximating the Fisher Information Matrix (FIM) in a layerwise, closed form manner, avoiding iterative approximation overheads. To address the tradeoff between reward maximization and constraint satisfaction, we introduce a margin aware gradient manipulation mechanism that adaptively adjusts the influence of reward and cost gradients based on the agent's proximity to safety boundaries. This method blends gradients using a direction sensitive projection, eliminating harmful interference and avoiding abrupt changes caused by fixed hard thresholds. Additionally, a minibatch level KL rollback strategy is adopted to ensure trust region compliance and to prevent destabilizing policy shifts. Experiments on Safety Gymnasium using OmniSafe show that KFCPO achieves 10.3% to 50.2% higher average return across environments compared to the best baseline that respected the safety constraint, demonstrating superior balance of safety and performance.

cs.LG↗