Language models can notice an impossible engineering problem yet still report it as solved
cs.AI, cs.CE, cs.CL
Submitted: 2026-10-05
Updated: 2026-10-05
Code: https://github.com/nbbllxx0/Paired-flawed-engineering-problems
Sources
- SoM-1K: A Thousand-Problem Benchmark Dataset for Strength of Materials
- ABench-Physics: Benchmarking Physical Reasoning in LLMs via High-Difficulty and Dynamic Physics Problems
- Mastering Olympiad-Level Physics with Artificial Intelligence
- FEABench: Evaluating Language Models on Multiphysics Reasoning Ability
- FEM-Bench: A Structured Scientific Reasoning Benchmark for Evaluating Code-Generating LLMs
- CADEngBench: It Looks Like CAD, but Does It Work? Evaluating Parametric Design, Assembly Reasoning, and Physics Simulation
- PDEAgent-Bench: A Multi-Metric, Multi-Library Benchmark for PDE Solver Generation
- HLE-Verified: A Systematic Verification and Structured Revision of Humanity's Last Exam
- FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI
- BrokenMath: A Benchmark for Sycophancy in Theorem Proving with LLMs
- Probing the Critical Point (CritPt) of AI Reasoning: a Frontier Physics Research Benchmark
- Evaluating Large Language Models Trained on Code
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