Safe Learning Under Irreversible Dynamics via Asking for Help
cs.LG, cs.AI
Submitted: 2025-02-19
Updated: 2026-09-11
Comments: Accepted to JMLR
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- TASRA: a Taxonomy and Analysis of Societal-Scale Risks from AI
- An Overview of Catastrophic AI Risks
- Delegative Reinforcement Learning: learning to avoid traps with a little help
- Regret Bounds for Discounted MDPs
- The AI risk repository: A meta-review, database, and taxonomy of risks from artificial intelligence
- Learning Adversarial MDPs with Stochastic Hard Constraints
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: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks