Read and Ignore: What Self-Directed LLM Agents Do in Public Service Scenarios
cs.AI, cs.HC, cs.MA
Submitted: 2026-11-11
Updated: 2026-11-11
Code: https://github.com/koizachek/selfdirected-llms
Terminology
Sources
- AgentHarm: A Benchmark for Measuring Harmfulness of LLM Agents
- Emergent Tool Use From Multi-Agent Autocurricula
- Leveraging Procedural Generation to Benchmark Reinforcement Learning
- AgentDojo: A Dynamic Environment to Evaluate Prompt Injection Attacks and Defenses for LLM Agents
- Go-Explore: a New Approach for Hard-Exploration Problems
- MONA: Myopic Optimization with Non-myopic Approval Can Mitigate Multi-step Reward Hacking
- Prioritized Level Replay
- Open-World Evaluations for Measuring Frontier AI Capabilities
- Scalable Evaluation of Multi-Agent Reinforcement Learning with Melting Pot
- What Makes LLM Agent Simulations Useful for Policy Practice? An Iterative Design Study in Emergency Preparedness
- AI and the Everything in the Whole Wide World Benchmark
- Goal Misgeneralization: Why Correct Specifications Aren't Enough For Correct Goals
- A Definition of Open-Ended Learning Problems for Goal-Conditioned Agents
- Voyager: An Open-Ended Embodied Agent with Large Language Models
- OdysseyBench: Evaluating LLM Agents on Long-Horizon Complex Office Application Workflows
- WebArena: A Realistic Web Environment for Building Autonomous Agents
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