Evaluating Model-Free Policy Optimization in Masked-Action Environments via an Exact Blackjack Oracle

• URL: http://arxiv.org/abs/2603.18642v1
• Date: Thu, 19 Mar 2026 09:08:59 GMT
• Title: Evaluating Model-Free Policy Optimization in Masked-Action Environments via an Exact Blackjack Oracle
• Abstract summary: An exact dynamic programming (DP) oracle was derived over 4,600 canonical decision cells.<n>It yielded ground-truth action values, optimal policy labels, and a theoretical expected value (EV) of -0.00161 per hand.<n>All methods exhibited substantial cell-conditional regret, indicating persistent policy-level errors.
• Score: 0.40611352512781873
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Infinite-shoe casino blackjack provides a rigorous, exactly verifiable benchmark for discrete stochastic control under dynamically masked actions. Under a fixed Vegas-style ruleset (S17, 3:2 payout, dealer peek, double on any two, double after split, resplit to four), an exact dynamic programming (DP) oracle was derived over 4,600 canonical decision cells. This oracle yielded ground-truth action values, optimal policy labels, and a theoretical expected value (EV) of -0.00161 per hand. To evaluate sample-efficient policy recovery, three model-free optimizers were trained via simulated interaction: masked REINFORCE with a per-cell exponential moving average baseline, simultaneous perturbation stochastic approximation (SPSA), and the cross-entropy method (CEM). REINFORCE was the most sample-efficient, achieving a 46.37% action-match rate and an EV of -0.04688 after 10^6 hands, outperforming CEM (39.46%, 7.5x10^6 evaluations) and SPSA (38.63%, 4.8x10^6 evaluations). However, all methods exhibited substantial cell-conditional regret, indicating persistent policy-level errors despite smooth reward convergence. This gap shows that tabular environments with severe state-visitation sparsity and dynamic action masking remain challenging, while aggregate reward curves can obscure critical local failures. As a negative control, it was proven and empirically confirmed that under i.i.d. draws without counting, optimal bet sizing collapses to the table minimum. In addition, larger wagers strictly increased volatility and ruin without improving expectation. These results highlight the need for exact oracles and negative controls to avoid mistaking stochastic variability for genuine algorithmic performance.

Related papers

• On-Policy Self-Distillation in Diffusion Models [89.32656931795293]
  Reinforcement learning can align diffusion models with human preferences and task-specific objectives, but endpoint rewards do not specify how an intermediate denoising prediction should change.<n>We introduce DiffusionOPSD as an on-policy self-distillation framework that converts image-level reward guidance into explicit targets for clean-output predictions at sampled queries.<n>Across SD 3.5-M and the step-distilled Z-Image-Turbo, our approach achieves the best final held-out scores in 19 of 20 reward-matched settings across two backbones and ten evaluators.
  arXiv  Detail & Related papers  (2026-08-25T14:55:24Z)
• Discrete Action Space as a Prerequisite for GRPO Convergence in Small-Model Continuous Control [0.0]
  Group Relative Policy Optimization can fine-tune small language models for simulated quadrotor continuous-control tasks.<n>In our benchmark, vanilla GRPO fine-tuning of Qwen-0.5B for 25 Hz quadrotor velocity control collapses to the trivial zero action.<n>A high-fidelity simulation using Crazyflie 2.1 dynamics surfaces a hover-region training-distribution gap.
  arXiv  Detail & Related papers  (2026-07-07T17:14:19Z)
• Reward Granularity in RLVR: Comparing Process and Outcome Reward Structures for Mathematical Reasoning in Small Language Models [0.0]
  Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as a promising paradigm for improving mathematical reasoning in language models.<n>We systematically compare five reward conditions applied to Qwen2.5-0.5B fine-tuned with Group Relative Policy Optimization (GRPO) on GSM8K.<n>Our results demonstrate that reward derivation is a first-order design decision for RLVR, with process-level supervision substantially improving both accuracy and trace fidelity in small language models.
  arXiv  Detail & Related papers  (2026-07-03T02:16:06Z)
• RVPO: Risk-Sensitive Alignment via Variance Regularization [13.192921543523283]
  We propose Reward-Variance Policy Optimization (RVPO), a risk-sensitive framework that penalizes inter-reward variance during advantage aggregation.<n>We show via Taylor expansion that a LogSumExp (SoftMin) operator effectively acts as a smooth variance penalty.<n>By preventing the model from neglecting difficult constraints to exploit easier objectives, RVPO improves overall scores on HealthBench.
  arXiv  Detail & Related papers  (2026-05-07T06:43:05Z)
• $V_{0.5}$: Generalist Value Model as a Prior for Sparse RL Rollouts [81.48669089692189]
  Generalist Value Models (such as $V_0.5$) achieve pre-trained value estimation by explicitly encoding model capabilities in-context.<n>We propose $V_0.5$, which adaptively fuses the baseline predicted by such value model with the empirical mean derived from sparse rollouts.<n>$V_0.5$ significantly outperforms GRPO and DAPO, achieving faster convergence and over some 10% performance improvement.
  arXiv  Detail & Related papers  (2026-03-11T14:57:41Z)
• Unbiased Dynamic Pruning for Efficient Group-Based Policy Optimization [60.87651283510059]
  Group Relative Policy Optimization (GRPO) effectively scales LLM reasoning but incurs prohibitive computational costs.<n>We propose Dynamic Pruning Policy Optimization (DPPO), a framework that enables dynamic pruning while preserving unbiased gradient estimation.<n>To mitigate the data sparsity induced by pruning, we introduce Dense Prompt Packing, a window-based greedy strategy.
  arXiv  Detail & Related papers  (2026-03-04T14:48:53Z)
• ODAR: Principled Adaptive Routing for LLM Reasoning via Active Inference [60.958331943869126]
  ODAR-Expert is an adaptive routing framework that optimize the accuracy-efficiency trade-off via principled resource allocation.<n>We show strong and consistent gains, including 98.2% accuracy on MATH and 54.8% on Humanity's Last Exam.
  arXiv  Detail & Related papers  (2026-02-27T05:22:01Z)
• Regularized Meta-Learning for Improved Generalization [0.0]
  Regularized meta-learning is a stable and deployment-efficient stacking strategy for high-dimensional ensemble systems.<n>A four-stage pipeline combines redundancy-aware projection, statistical meta-feature augmentation, and cross-validated regularized meta-models.<n>On the Playground Series S6E1 benchmark, the proposed framework achieves an out-of-fold RMSE of 8.582, improving over simple averaging (8.894) and conventional Ridge stacking (8.627), while matching greedy hill climbing (8.603) with substantially lower runtime (4 times faster)
  arXiv  Detail & Related papers  (2026-02-12T22:55:32Z)
• ACAR: Adaptive Complexity Routing for Multi-Model Ensembles with Auditable Decision Traces [3.151184728006369]
  We present ACAR, a measurement framework for studying multi-model orchestration under auditable conditions.<n>ACAR uses self-consistency variance (sigma) computed from N=3 probe samples to route tasks across single-model, two-model, and three-model execution modes.<n>We evaluate ACAR on 1,510 tasks spanning four benchmarks, producing more than 7,550 auditable runs.
  arXiv  Detail & Related papers  (2026-02-06T23:27:17Z)
• Prompt Injection Evaluations: Refusal Boundary Instability and Artifact-Dependent Compliance in GPT-4-Series Models [0.0]
  We evaluate two models, GPT-4.1 and GPT-4o, using 3,274 perturbation runs derived from refusal-inducing prompt injection attempts.<n>We find that artifact type is a stronger predictor of refusal failure than perturbation style.<n>We conclude that refusal behavior is a probabilistic, artifact-dependent boundary phenomenon rather than a stable binary property.
  arXiv  Detail & Related papers  (2026-01-25T17:14:33Z)
• Non-Convex Portfolio Optimization via Energy-Based Models: A Comparative Analysis Using the Thermodynamic HypergRaphical Model Library (THRML) for Index Tracking [0.0]
  This paper introduces a novel approach using THRML (Thermodynamic HypergRaphical Model Library), a JAX-based library for building and sampling probabilistic graphical models.<n> THRML samples from the Boltzmann distribution of high-quality portfolios using GPU-accelerated block Gibbs sampling, providing natural regularization against overfitting.<n>Backtesting on a 100-stock S and P 500 universe from 2023 to 2025 demonstrates that THRML achieves 4.31 percent annualized tracking error versus 5.66 to 6.30 percent for baselines, while simultaneously generating 128.63 percent total return.
  arXiv  Detail & Related papers  (2026-01-12T18:04:33Z)
• Optimistic Feasible Search for Closed-Loop Fair Threshold Decision-Making [0.0]
  We study online learning of a one-dimensional threshold policy from bandit feedback.<n>We propose Optimistic Feasible Search (OFS), a simple grid-based method that maintains confidence bounds for reward and constraint residuals.
  arXiv  Detail & Related papers  (2025-12-26T10:44:40Z)
• From Noisy Traces to Stable Gradients: Bias-Variance Optimized Preference Optimization for Aligning Large Reasoning Models [90.45197506653341]
  Large reasoning models generate intermediate reasoning traces before producing final answers.<n> aligning LRMs with human preferences, a crucial prerequisite for model deployment, remains underexplored.<n>A common workaround optimized a single sampled trajectory, which introduces substantial gradient variance from trace sampling.
  arXiv  Detail & Related papers  (2025-10-06T17:58:01Z)
• Squeeze the Soaked Sponge: Efficient Off-policy Reinforcement Finetuning for Large Language Model [56.92219181993453]
  We propose Reincarnating Mix-policy Proximal Policy Gradient (ReMix) to enable on-policyRFT methods like PPO and GRPO to leverage off-policy data.<n>ReMix consists of three major components: (1) Mix-policy proximal policy gradient with an increased Update-To-Data (UTD) ratio for efficient training; (2) KL-Convex policy constraint to balance the trade-off between stability and flexibility; (3) Policy reincarnation to achieve a seamless transition from efficient early-stage learning to steady improvement.
  arXiv  Detail & Related papers  (2025-07-09T14:29:45Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.