How Do Agent Harnesses Create Value? Planning Information and Release Control in Stateful LLM Agents
cs.AI
Submitted: 2026-09-17
Updated: 2026-09-17
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- From Confident Closing to Silent Failure: Characterizing False Success in LLM Agents
- Large Language Monkeys: Scaling Inference Compute with Repeated Sampling
- Beyond Task Completion: Revealing Corrupt Success in LLM Agents through Procedure-Aware Evaluation
- Evaluating Large Language Models Trained on Code
- Done, But Not Sure: Disentangling World Completion from Self-Termination in Embodied Agents
- Training Verifiers to Solve Math Word Problems
- MemGPT: Towards LLMs as Operating Systems
- Agent Planning Benchmark: A Diagnostic Framework for Planning Capabilities in LLM Agents
- Solving math word problems with process- and outcome-based feedback
- On the Brittle Foundations of ReAct Prompting for Agentic Large Language Models
- AI Harness Engineering: A Runtime Substrate for Foundation-Model Software Agents
Related papers
- 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
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
- VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection