Agent Behavior as Code: Efficient and Robust LLM Agents with Programmatic Specifications
cs.AI, cs.CL, cs.MA
Submitted: 2026-10-04
Updated: 2026-10-04
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- Neural Module Networks
- $\tau^2$-Bench: Evaluating Conversational Agents in a Dual-Control Environment
- WorkArena++: Towards Compositional Planning and Reasoning-based Common Knowledge Work Tasks
- Evaluating Large Language Models Trained on Code
- Metamorphic Testing: A New Approach for Generating Next Test Cases
- Program of Thoughts Prompting: Disentangling Computation from Reasoning for Numerical Reasoning Tasks
- DreamCoder: Growing generalizable, interpretable knowledge with wake-sleep Bayesian program learning
- LILO: Learning Interpretable Libraries by Compressing and Documenting Code
- Putnam-AXIOM: A Functional and Static Benchmark for Measuring Higher Level Mathematical Reasoning in LLMs
- Automated Design of Agentic Systems
- Deep Research Agents: A Systematic Examination And Roadmap
- DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines
- Generalization without systematicity: On the compositional skills of sequence-to-sequence recurrent networks
- Code as Policies: Language Model Programs for Embodied Control
- GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models
- Shared Selective Persistent Memory for Agentic LLM Systems
- Neural Programmer-Interpreters
- Toolformer: Language Models Can Teach Themselves to Use Tools
- ReGAL: Refactoring Programs to Discover Generalizable Abstractions
- Voyager: An Open-Ended Embodied Agent with Large Language Models
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
- Agents' Last Exam
- 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