Iterative Policy Refinement through Semantic Rollout Analysis
cs.LG, cs.AI
Submitted: 2026-10-01
Updated: 2026-10-01
Sources
- Weird Generalization and Inductive Backdoors: New Ways to Corrupt LLMs
- GaP: A Graph-as-Policy Multi-Agent Self-Learning Harness For Variational Automation Tasks
- ELEMENTAL: Interactive Learning from Demonstrations and Vision-Language Models for Reward Design in Robotics
- Teaching Large Language Models to Self-Debug
- MoDem: Accelerating Visual Model-Based Reinforcement Learning with Demonstrations
- VoxPoser: Composable 3D Value Maps for Robotic Manipulation with Language Models
- Interactive Program Synthesis for Modeling Collaborative Physical Activities from Narrated Demonstrations
- Code as Policies: Language Model Programs for Embodied Control
- LLM+P: Empowering Large Language Models with Optimal Planning Proficiency
- DARTS: Differentiable Architecture Search
- ASPIRE: Agentic /Skills Discovery for Robotics
- Measuring Competency, Not Performance: Item-Aware Evaluation Across Medical Benchmarks
- Eureka: Human-Level Reward Design via Coding Large Language Models
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
- An Algorithmic Perspective on Imitation Learning
- Adaptive Language-Guided Abstraction from Contrastive Explanations
- Learning with Language-Guided State Abstractions
- ENPIRE: Agentic Robot Policy Self-Improvement in the Real World
- Darwin Godel Machine: Open-Ended Evolution of Self-Improving Agents
- Group Sequence Policy Optimization
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks