Found by Search, Broken by a Shortcut: Transfer and Tampering of Searched Intrinsic Rewards
cs.LG, cs.AI
Submitted: 2026-11-03
Updated: 2026-11-03
Code: https://github.com/chasemetoyer/found-by-search
Terminology
Sources
- Meta-learning curiosity algorithms
- Concrete Problems in AI Safety
- Unifying Count-Based Exploration and Intrinsic Motivation
- Large-Scale Study of Curiosity-Driven Learning
- Exploration by Random Network Distillation
- Minigrid & Miniworld: Modular & Customizable Reinforcement Learning Environments for Goal-Oriented Tasks
- Reward Tampering Problems and Solutions in Reinforcement Learning: A Causal Influence Diagram Perspective
- The CMA Evolution Strategy: A Tutorial
- CleanRL: High-quality Single-file Implementations of Deep Reinforcement Learning Algorithms
- Eureka: Human-Level Reward Design via Coding Large Language Models
- Satisfied, Not Solving: Wireheading of Bounded Homeostatic Rewards in Reinforcement Learning
- Discovering Reinforcement Learning Algorithms
- RIDE: Rewarding Impact-Driven Exploration for Procedurally-Generated Environments
- Proximal Policy Optimization Algorithms
- Defining and Characterizing Reward Hacking
- Meta-Gradient Reinforcement Learning
- On Learning Intrinsic Rewards for Policy Gradient Methods
- What Can Learned Intrinsic Rewards Capture?
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis