# CodeCrash: Exposing LLM Fragility to Misleading Natural Language in Code Reasoning

arXiv: https://arxiv.org/abs/2504.14119 · HTML: https://mlsanity.com/paper/2504.14119/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning.
- Builds on 31 of its 65 references that are in the map.
- First posted on arXiv 19 Apr 2025.

## Placed in

- LLM reasoning / [Reasoning for agents, planning and code](https://mlsanity.com/lr/reasoning-for-agents-planning-and-code/)

## Builds on (most cited first, 30 of 31)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning](https://mlsanity.com/paper/2501.12948/) (2025)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [GPT-4o System Card](https://mlsanity.com/paper/2410.21276/) (2024)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [Qwen2.5-Coder Technical Report](https://mlsanity.com/paper/2409.12186/) (2024)
- [Measuring Coding Challenge Competence With APPS](https://mlsanity.com/paper/2105.09938/) (2021)
- [Show Your Work: Scratchpads for Intermediate Computation with Language Models](https://mlsanity.com/paper/2112.00114/) (2021)
- [DeepSeek-V3 Technical Report](https://mlsanity.com/paper/2412.19437/) (2024)
- [Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs](https://mlsanity.com/paper/2412.21187/) (2024)
- [The Danger of Overthinking: Examining the Reasoning-Action Dilemma in Agentic Tasks](https://mlsanity.com/paper/2502.08235/) (2025)
- [Large Language Models Meet NL2Code: A Survey](https://mlsanity.com/paper/2212.09420/) (2022)
- [ReCode: Robustness Evaluation of Code Generation Models](https://mlsanity.com/paper/2212.10264/) (2022)
- [OverThink: Slowdown Attacks on Reasoning LLMs](https://mlsanity.com/paper/2502.02542/) (2025)
- [CodeScore: Evaluating Code Generation by Learning Code Execution](https://mlsanity.com/paper/2301.09043/) (2023)
- [On the Robustness of Code Generation Techniques: An Empirical Study on GitHub Copilot](https://mlsanity.com/paper/2302.00438/) (2023)
- [NExT: Teaching Large Language Models to Reason about Code Execution](https://mlsanity.com/paper/2404.14662/) (2024)
- [Interaction2Code: Benchmarking MLLM-based Interactive Webpage Code Generation from Interactive Prototyping](https://mlsanity.com/paper/2411.03292/) (2024)
- [Reasoning Runtime Behavior of a Program with LLM: How Far Are We?](https://mlsanity.com/paper/2403.16437/) (2024)
- [On Robustness of Prompt-based Semantic Parsing with Large Pre-trained Language Model: An Empirical Study on Codex](https://mlsanity.com/paper/2301.12868/) (2023)
- [StructCoder: Structure-Aware Transformer for Code Generation](https://mlsanity.com/paper/2206.05239/) (2022)
- [Code Execution with Pre-trained Language Models](https://mlsanity.com/paper/2305.05383/) (2023)
- [Reinforcement Learning from Automatic Feedback for High-Quality Unit Test Generation](https://mlsanity.com/paper/2310.02368/) (2023)
- [NLPerturbator: Studying the Robustness of Code LLMs to Natural Language Variations](https://mlsanity.com/paper/2406.19783/) (2024)
- [Transforming Software Development: Evaluating the Efficiency and Challenges of GitHub Copilot in Real-World Projects](https://mlsanity.com/paper/2406.17910/) (2024)
- [Codev-Bench: How Do LLMs Understand Developer-Centric Code Completion?](https://mlsanity.com/paper/2410.01353/) (2024)
- [MdEval: Massively Multilingual Code Debugging](https://mlsanity.com/paper/2411.02310/) (2024)
- [Large Language Models as Code Executors: An Exploratory Study](https://mlsanity.com/paper/2410.06667/) (2024)

## Built on by (most cited first, 14 of 14)

- [DesignBench: A Comprehensive Benchmark for MLLM-based Front-end Code Generation](https://mlsanity.com/paper/2506.06251/) (2025)
- [A Comprehensive Survey on Trustworthiness in Reasoning with Large Language Models](https://mlsanity.com/paper/2509.03871/) (2025)
- [CoRe: Benchmarking LLMs Code Reasoning Capabilities through Static Analysis Tasks](https://mlsanity.com/paper/2507.05269/) (2025)
- [SWE-Chain: Benchmarking Coding Agents on Chained Release-Level Package Upgrades](https://mlsanity.com/paper/2605.14415/) (2026)
- [IRG-MotionLLM: Interleaving Motion Generation, Assessment and Refinement for Text-to-Motion Generation](https://mlsanity.com/paper/2512.10730/) (2025)
- [Conflict-Aware Fusion: Mitigating Logic Inertia in Large Language Models via Structured Cognitive Priors](https://mlsanity.com/paper/2512.06393/) (2025)
- [Aletheia: What Makes RLVR For Code Verifiers Tick?](https://mlsanity.com/paper/2601.12186/) (2026)
- [Refining Critical Thinking in LLM Code Generation: A Faulty Premise-based Evaluation Framework](https://mlsanity.com/paper/2508.03622/) (2025)
- [TREAT: A Code LLMs Trustworthiness / Reliability Evaluation and Testing Framework](https://mlsanity.com/paper/2510.17163/) (2025)
- [Themis: Training Robust Multilingual Code Reward Models for Flexible Multi-Criteria Scoring](https://mlsanity.com/paper/2605.00754/) (2026)
- [The Invisible Lottery: How Subtle Cues Steer Algorithm Choice in LLM Code Generation](https://mlsanity.com/paper/2606.04057/) (2026)
- [Towards Evaluation of Implicit Software World Models in Coding LLMs](https://mlsanity.com/paper/2606.27406/) (2026)
- [Wrong Design Intent Is Worse Than Never Conditioning: A Derangement-Control Diagnosis of Header Conditioning in CAD Program Completion](https://mlsanity.com/paper/2607.23191/) (2026)
- [ALIBI: Adaptive Agentic Attacks on LLM-Based Vulnerability Detectors via Adversarial Code Comments](https://mlsanity.com/paper/2607.24964/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) have recently demonstrated strong capabilities in code-related tasks, but their robustness in code reasoning under perturbations remains underexplored. We introduce CodeCrash, a stress-testing framework with 1,279 questions from CruxEval and LiveCodeBench, designed to evaluate reasoning reliability under structural perturbations and misleading natural language (NL) contexts. Through a systematic evaluation of 17 LLMs, we find that models often shortcut reasoning by over-relying on NL cues, leading to an average performance degradation of 23.2% in output prediction tasks. Even with Chain-of-Thought reasoning, models on average still have a 13.8% drop due to distractibility and rationalization, revealing a lack of critical reasoning capability to distinguish the actual code behaviors. While Large Reasoning Models with internal reasoning mechanisms improve robustness by fostering critical thinking, plausible yet incorrect hints can trigger pathological self-reflection, causing 2-3 times token consumption and even catastrophic cognitive dissonance in extreme cases for QwQ-32B. We refer to this phenomenon as Reasoning Collapse. CodeCrash provides a rigorous benchmark for evaluating robustness in code reasoning, guiding future research and development toward more reliable and resilient models.