This paper applies CPI to deep RL, improving stability and performance.
problem Improving stability and performance in deep reinforcement learning.
method Combines Conservative Policy Iteration with deep neural networks and adaptive mixture rates.
result Demonstrates improved stability and performance in deep RL algorithms.
Theoretical analysis confirms non-conservative algorithms can converge to optimal policies.
problem Theoretical guarantees for non-conservative reinforcement learning algorithms.
method Theoretical analysis of Peng's Q( λ λ λ ) algorithm. result Peng's Q( λ λ λ ) converges to an optimal policy under certain conditions. BCPO optimizes offline RL policies by converting uncertainty into conservative bounds.
problem Offline RL's fragility under distribution shifts and model errors.
method Bayesian approach with credible lower bounds and KL regularization.
result BCPO yields an uncertainty-calibrated policy that avoids exploiting model errors.
Combines soft greediness with Modified Policy Iteration for more efficient deep reinforcement learning.
problem Improving sample efficiency in deep reinforcement learning.
method Combines soft greediness with Modified Policy Iteration (MPI) for off-policy learning.
result The proposed algorithm is more sample efficient than original PPO and competitive with SAC.
The paper tackles safe exploration in RL by a conservative safety critic.
problem Safe exploration in reinforcement learning (RL) when partially trained policies are deployed.
method Learning a conservative safety estimate through a critic, provably bounding catastrophic failures.
result The approach provably converges to competitive task performance with significantly lower catastrophic failure rates.
Paper analyzes convergence of entropy-regularized reinforcement learning methods.
problem Ensuring sub-optimality control in entropy-regularized RL.
method Unified analysis of regularized MPI and VI schemes, providing convergence rates.
result Established sufficient conditions for convergence and provided explicit rates.
Algorithm corrects mistakes in imitation learning tasks.
problem Covariate shift problem in learning from demonstrations.
method Value Iteration with Negative Sampling (VINS) for conservative extrapolation.
result VINS corrects mistakes of the behavioral cloning policy.
Paper connects DP and optimization for RL, suggesting new algorithms.
problem Optimizing scalar objectives in RL.
method Drawing connections between DP and optimization algorithms.
result Links between DP schemes and optimization algorithms.
Proposes a conservative exploration method for RL agents.
problem Guaranteeing performance of exploratory policies in RL.
method Importance sampling for off-policy policy evaluation.
result Derives a regret bound ensuring no conservative constraint violation.
We consider the infinite-horizon discounted optimal control problem formalized by Markov Decision Processes. We focus on several approximate variations of the Policy Iteration algorithm: Approximate Policy Iteration, Conservative Policy Iteration (CPI), a natural adaptation of the Policy Search by Dynamic Programming a…
Auto-CEI improves LLM reasoning by balancing assertiveness and conservativeness.
problem Hallucinations and laziness in LLM reasoning tasks.
method Expert Iteration explores reasoning trajectories, guiding incorrect paths back on track and promoting appropriate 'I don't know' responses.
result Auto-CEI achieves superior alignment in logical reasoning, mathematics, and planning tasks.
Real-world applications require RL algorithms to act safely. During learning process, it is likely that the agent executes sub-optimal actions that may lead to unsafe/poor states of the system. Exploration is particularly brittle in high-dimensional state/action space due to increased number of low-performing actions. …
A new method learns robust policies from offline data with latent structures.
problem Conservative policies under unrealistic dynamics shifts.
method d-RRMDP framework with f f f -divergence regularization and R2PVI algorithm. result R2PVI learns robust policies with superior computational efficiency.
TRPO adapts trust region methods for faster RL convergence.
problem Improving RL convergence rates in regularized MDPs.
method Adaptive scaling in TRPO for faster convergence rates.
result First RL result showing faster rates with regularization.
HAMBO estimates policy performance by hallucinating worst-case trajectories, providing valid lower bounds.
problem Conservative off-policy evaluation of policies in real-world applications.
method HAMBO hallucinates worst-case trajectories based on learned model uncertainty.
result Valid lower bounds on policy performance, converging to true expected return under regular conditions.
New algorithms ensure policies perform at least as good as a baseline in reinforcement learning.
problem Learning policies that are guaranteed to perform at least as well as a baseline in reinforcement learning.
method Introduce conservative exploration for average reward and finite horizon problems, presenting two optimistic algorithms.
result Guaranteed performance of policies at least as good as a baseline, without hindering learning ability.
Paper proposes an RL algorithm to ensure policy performance guarantees.
problem Lack of performance guarantees for RL policies compared to baselines.
method Online model-free algorithm that ensures conservative exploration.
result Regret bound of i l d e O ( T ) ilde{\mathcal{O}}(\sqrt{T}) i l d e O ( T ) for both discrete and continuous spaces. StepMix algorithm ensures safe exploration in reinforcement learning with near-optimal performance.
problem Conservative exploration in reinforcement learning under episode-wise constraints.
method StepMix algorithm balances exploitation and exploration while ensuring episode-wise conservative constraint.
result StepMix achieves near-optimal regret order as in the constraint-free setting.
CQL learns conservative Q-functions to improve offline RL performance.
problem Leveraging large, static datasets in reinforcement learning without further interaction.
method Conservative Q-learning (CQL) which learns a conservative Q-function to lower-bound policy values.
result CQL substantially outperforms existing offline RL methods, often achieving 2-5 times higher final returns.
VaR-CPO optimizes VaR-constrained RL problems with conservative policy updates.
problem Optimizing VaR-constrained reinforcement learning problems.
method Combines Cantelli's inequality and trust-region framework for efficient and conservative optimization.
result Achieves zero constraint violations during training in feasible environments.
SCPO learns robust policies without modeling disturbance, improving real-world task performance.
problem Poor performance of reinforcement learning in real-world tasks due to disturbance in transition dynamics.
method State-conservative policy optimization (SCPO) that reduces disturbance to state space and approximates it with a gradient-based regularizer.
result SCPO learns robust policies without prior knowledge of disturbance or simulators, improving performance in robot control tasks.
Canary optimizes VaR-constrained RL problems with a conservative bound using Cantelli's inequality.
problem Optimizing reinforcement learning policies under VaR constraints in dense cost regimes.
method Employing Cantelli's inequality to create a conservative and smooth bound on VaR constraints based on moments of cost returns. Extending trust-region framework for worst-case bounds on policy improvement and constraint violation.
result Canary reliably satisfies VaR constraints with fewest violations and earliest permanent satisfaction, while maintaining reward competitiveness.
RORL improves offline RL robustness with conservative smoothing.
problem Distribution shift and robustness issues in offline RL.
method RORL introduces regularization and conservative smoothing for robustness.
result RORL achieves state-of-the-art performance and robustness to adversarial perturbations.
New algorithm improves online learning under performance constraints.
problem Improving performance of existing systems in various fields.
method Conservative Constrained LinUCB (CLUCB2) algorithm for contextual linear bandits.
result Empirically outperforms existing conservative bandit algorithms.
MoMA improves model-based RL by using unrestricted policy classes.
problem Limited sample efficiency and generalizability in model-based offline RL.
method Model-based mirror ascent algorithm with general function approximations.
result Theoretical guarantees and practical implementation of MoMA.
New RL algorithms learn policies competitive with best in class without assuming optimal policy.
problem Agnostic RL where best policy not necessarily optimal.
method First-order optimization in non-Euclidean space, reducing to policy learning.
result Sample complexity upper bounds for three algorithms under VGD condition.
New framework models non-conservative stochastic processes without energy conservation constraints.
problem Existing Schrödinger Bridge methods are limited by energy-conservation assumptions.
method Introduces non-conservative generalized Schrödinger bridge (NCGSB) based on contact Hamiltonian mechanics.
result Contact Wasserstein geodesic (CWG) provides a broader class of real-world stochastic processes.
New batch RL method avoids overly optimistic policies.
problem Challenges in applying RL to large domains with limited data.
method Modified Bellman optimality and evaluation back-up for conservative updates.
result Can find approximately best policies within explored state-action space.
Paper presents a reduction-based framework for conservative bandits and RL with improved lower and upper bounds.
problem Conservative bandits and reinforcement learning problems.
method Reduction technique to calculate necessary and sufficient budget from baseline policy.
result Improved lower and upper bounds for various conservative settings.
Robust Reinforcement Learning aims to derive optimal behavior that accounts for model uncertainty in dynamical systems. However, previous studies have shown that by considering the worst case scenario, robust policies can be overly conservative. Our soft-robust framework is an attempt to overcome this issue. In this pa…
Proposes a method to avoid excessive exploration in reinforcement learning.
problem Avoiding excessive exploration in reinforcement learning to deploy it in practice.
method Designs a novel algorithm using UCB reinforcement learning policy with adaptive exploration constraints.
result Proves that the approach remains conservative while minimizing regret in tabular settings and validates on real-world tasks.
DSPI connects natural policy gradient to policy iteration, proving global convergence.
problem Optimizing policies in reinforcement learning.
method DSPI framework, combining smoothed policy iteration and natural policy gradient.
result DSPI achieves geometric convergence and optimal complexity for policy optimization.
Offline RL policies should adapt to unknown aspects of the environment.
problem Uncertainty in offline RL datasets leads to suboptimal policies.
method Adaptive policies that consider all transitions seen so far, solving an implicit POMDP.
result Optimal adaptive policies improve offline RL performance.
Higher conservative training increases reward-hacking in reasoning models.
problem Reward hacking during online adaptation in reasoning models.
method Conservative offline training with varying levels of conservatism (β) was applied to a Qwen3-14B policy, and online adaptation was measured against a reward ensemble.
result Higher conservatism (β) increases reward-hacking damage, measured by the Goodhart gap and AUGC.
POMBU improves model-based RL's asymptotic performance by estimating and using uncertainty.
problem Model-based reinforcement learning struggles with model errors, leading to suboptimal performance.
method POMBU uses estimated uncertainty to optimize policies conservatively, improving asymptotic performance.
result POMBU outperforms existing methods in sample efficiency and asymptotic performance.
Enhances deep RL learning with stable policy updates.
problem Stability and performance degradation in deep RL methods.
method EVEREST method providing conservative updates with confidence bounds.
result Significant improvements in continuous control and Atari benchmarks.
Paper proves exponential lower bounds for policy iteration in MDPs.
problem Proving lower bounds for policy iteration in multi-action MDPs.
method Generalized previous results to k-action MDPs, constructed families of MDPs.
result Proved novel exponential lower bound of (3+k)2^(N/2-3) iterations.
Proposes a method to improve few-shot transfer in off-dynamics RL.
problem Traditional RL struggles with transferring policies between environments with different dynamics.
method Introduces a penalty to regulate source-trained policies in target environments with limited data.
result Improves performance in various off-dynamics RL scenarios compared to existing methods.
Policy gradient methods with aggregated states can achieve better performance than approximate policy iteration.
problem Approximation errors in policy and value function approximations.
method State-aggregated representations and policy gradient methods.
result Policy gradient methods can achieve a per-period regret bounded by ε, while approximate policy iteration and value iteration have a higher regret.
Adaptive optimal control of nonlinear dynamic systems with deterministic and known dynamics under a known undiscounted infinite-horizon cost function is investigated. Policy iteration scheme initiated using a stabilizing initial control is analyzed in solving the problem. The convergence of the iterations and the optim…
New algorithms optimize a soft-robust criterion in reinforcement learning, reducing conservatism.
problem Computing robust policies for high-stakes decisions with limited data.
method Soft-robust criterion using risk measures, two algorithms for optimization.
result Our algorithms produce less conservative solutions than existing methods.
One-step policy improvement outperforms iterative RL methods on D4RL.
problem Improving offline RL without off-policy evaluation.
method One-step constrained/regularized policy improvement using on-policy Q estimates.
result One-step algorithm outperforms iterative algorithms on D4RL benchmark.
A gauged bi-differential calculus over an associative (and not necessarily commutative) algebra A is an N-graded left A-module with two covariant derivatives acting on it which, as a consequence of certain (e.g., nonlinear differential) equations, are flat and anticommute. As a consequence, there is an iterative constr…
Develops a support-aware framework for reserve-policy selection in advertising markets.
problem Log-based reserve-price evaluation risks weak support and subgroup harm.
method Support-aware offline decision framework converting logged evidence into certified policies.
result Preserves the best gate-passing policy while eliminating only policies with certified regret.
Develops algorithms for CCBs with non-linear costs, improving safety and performance.
problem Safety constraints in sequential decision making with non-linear arm costs.
method Innovative algorithms using Inverse Gap Weighting (IGW) and online regression oracle.
result Sub-linear regret bounds for C-SquareCB and first-order regret for C-FastCB.
RRPI improves offline RL by optimizing policies against worst-case dynamics.
problem Offline RL's performance degrades under distribution shift and transition uncertainty.
method Formulates offline RL as robust policy optimization, treating transition kernel as decision variable.
result RRPI achieves strong average performance on D4RL benchmarks, outperforming recent baselines.
Paper proposes a new DRL algorithm optimizing Spectral Risk Measures for better risk management.
problem Inconsistencies and conservatism in existing risk measures in DRL.
method Optimizes a broader class of static Spectral Risk Measures (SRM) in DRL.
result Demonstrates improved performance over existing risk-neutral and risk-sensitive DRL models.
The paper proposes an iterative approach to batch reinforcement learning for safer and more informative data collection.
problem Learning policies that are too rigid and do not adapt to new data.
method Safe diversified model-based policy search in an iterative batch reinforcement learning framework.
result Improved learned policies through continuous data collection and adaptation.