DuET: Dual Execution for Test Output Prediction with Generated Code and Pseudocode

TL;DR

DuET is a dual-execution framework that combines code execution and pseudocode reasoning via LLMs to improve test output prediction accuracy, achieving state-of-the-art results.

Contribution

It introduces a novel dual-execution approach that leverages both code and pseudocode grounding, enhancing reliability in test output prediction.

Findings

DuET improves Pass@1 by 13.6 percentage points on LiveCodeBench.

Combining code execution and pseudocode reasoning addresses limitations of each approach.

Dual execution with majority voting outperforms individual methods.

Abstract

This work addresses test output prediction, a key challenge in test case generation. To improve the reliability of predicted outputs by LLMs, prior approaches generate code first to ground predictions. One grounding strategy is direct execution of generated code, but even minor errors can cause failures. To address this, we introduce LLM-based pseudocode execution, which grounds prediction on more error-resilient pseudocode and simulates execution via LLM reasoning. We further propose DuET, a dual-execution framework that combines both approaches by functional majority voting. Our analysis shows the two approaches are complementary in overcoming the limitations of direct execution suffering from code errors, and pseudocode reasoning from hallucination. On LiveCodeBench, DuET achieves the state-of-the-art performance, improving Pass@1 by 13.6 pp.