Rules to Tools: Executable Checks for LLM Agents in Scientific Computing
cs.AI, cs.SE
Submitted: 2026-09-29
Updated: 2026-09-29
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Sources
- LLM Assisted Coding with Metamorphic Specification Mutation Agent
- MDArena: Evaluating Coding Agents on Realistic Molecular Dynamics Workflows
- CodeSpecBench: Benchmarking LLMs for Executable Behavioral Specification Generation
- Rethinking the Value of Agent-Generated Tests for LLM-Based Software Engineering Agents
- AInsteinBench: Benchmarking Coding Agents on Scientific Repositories
- CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing
- PDEAgent-Bench: A Multi-Metric, Multi-Library Benchmark for PDE Solver Generation
- Lang-PINN: From Language to Physics-Informed Neural Networks via a Multi-Agent Framework
- SciCode-Verified: How Benchmark Defects Underestimated the Scientific-Coding Ability of Language Models
- Teaching Code LLMs to Reason with Intermediate Formal Specifications
- CodePDE: An Inference Framework for LLM-driven PDE Solver Generation
- Security Tests as Executable Specifications for LLM Code Generation: Benefits, Trade-offs, and Coverage Limits
- Self-Refine: Iterative Refinement with Self-Feedback
- Auto Research with Specialist Agents Develops Effective and Non-Trivial Training Recipes
- Auto Research for Materials: Auditable AI-Scientist Workflows with Held-Out Transfer
- The Implementation Lottery: Auditing Idea Reliability in Automated Research
- scicode-lint: Detecting Methodology Bugs in Scientific Python Code with LLM-Generated Patterns
- Reflexion: Language Agents with Verbal Reinforcement Learning
- SciCode: A Research Coding Benchmark Curated by Scientists
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