This paper provides a mathematical framework for studying different policy learning problems and shows reductions between them.

The paper introduces Prompted Policy Optimization (PromptPO), an LLM-based method that successfully optimizes policies for various sequential RL tasks, demonstrating that LLMs can replace classical RL…

The paper proposes Detector Evasion Policy Optimization (DEPO), a constrained reinforcement learning method that effectively evades AI text detectors while strictly maintaining the original text's sem…

The paper introduces Latent Policy Guardrail (LPG), a novel framework that efficiently enforces dynamic safety policies for LLMs by compressing complex policy deliberation into a small set of latent t…

The paper introduces a novel LLM-driven evolutionary framework to synthesize admissible, domain-specific pattern generators, enabling optimal classical planning with high performance and interpretabil…

HarnessForge introduces a meta-adaptive framework that jointly evolves the execution structure (harness) and the reasoning policy of LLM agents, significantly improving overall system performance acro…

This paper develops a policy-learning framework to optimally assign prediction tasks to multiple agents, considering individual agent expertise and capacity constraints, achieving systematic performan…

The paper introduces WIRE, a pipeline for diagnosing live intra-policy rule conflicts in LLM agents by identifying and testing specific rule pairs within a single prompt policy that can co-govern a re…

The paper introduces a learned 'rerooter' mechanism to improve subgoal-based policy tree search, allowing scalable search in complex environments without the overhead of explicit subgoal generation.

The OISD framework improves language model reasoning by distilling on-policy predictive signals from the final output layer to intermediate representations, leading to substantial improvements on math…

The paper proposes S2L-PO, a framework that uses smaller, naturally diverse models as structured explorers to enhance the policy-level diversity and performance of larger language models during traini…

Taiji is a novel LLM-as-Enhancer framework that optimizes recommender systems by addressing the challenges of generating high-quality reasoning data and balancing semantic and ID-based rewards.

S-SPPO introduces a dual-space semantic calibration framework to stabilize Self-Play Preference Optimization (SPPO), preventing policy degeneration when preference oracles assign overly confident wins…

The paper introduces Q-ALIGN DT, a novel framework that improves conditioned sequence models by enforcing alignment between the input return-to-go (RTG) signal and the output policy's expected Q-value…

ESPO is a novel reinforcement learning algorithm that detects trajectory failure in large language models and terminates rollouts early, significantly improving performance on mathematical reasoning b…

This paper develops a new algorithm for policy optimization in online episodic tabular Markov decision processes with unknown transition kernels, providing data-dependent regret bounds and best-of-bot…

This paper introduces the first LLM-generated, domain-independent heuristics for symbolic AI planning, using evolutionary search to surpass the performance of hand-engineered state-of-the-art methods.

The paper introduces Safe Equilibrium Policy Optimization (σepo{}) to train language models for multi-agent strategic tasks, achieving improved safety and robustness across various game domains.

The paper proposes a hybrid reasoning framework where Large Language Models (LLMs) generate code to encode complex optimization problems into a preference-based Maximum Satisfiability (MaxSAT) format,…