Sampling Is Not Curiosity: Why LLM Agents Should Investigate
cs.AI
Submitted: 2026-11-07
Updated: 2026-11-07
Terminology
Sources
- What Does It Take to Be a Good AI Research Agent? Studying the Role of Ideation Diversity
- Large-Scale Study of Curiosity-Driven Learning
- CDE: Curiosity-Driven Exploration for Efficient Reinforcement Learning in Large Language Models
- Can In-Context Learning Support Intrinsic Curiosity?
- Agents Explore but Agents Ignore: LLMs Lack Environmental Curiosity
- World Models
- MLAgentBench: Evaluating Language Agents on Machine Learning Experimentation
- Towards Continual Knowledge Learning of Language Models
- Meta-RL Induces Exploration in Language Agents
- In-context Reinforcement Learning with Algorithm Distillation
- WorldLLM: Improving LLMs' world modeling using curiosity-driven theory-making
- The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
- In-context Learning and Induction Heads
- CORAL: Towards Autonomous Multi-Agent Evolution for Open-Ended Discovery
- Voyager: An Open-Ended Embodied Agent with Large Language Models
- An Explanation of In-context Learning as Implicit Bayesian Inference
- The AI Scientist-v2: Workshop-Level Automated Scientific Discovery via Agentic Tree Search
- OMNI: Open-endedness via Models of human Notions of Interestingness
- Consistent Paths Lead to Truth: Self-Rewarding Reinforcement Learning for LLM Reasoning
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents