No Test Cases, No Problem: Distillation-Driven Code Generation for Scientific Workflows

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Raghavan, Siddeshwar, Mallick, Tanwi
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866918467657007104
author Raghavan, Siddeshwar
Mallick, Tanwi
author_facet Raghavan, Siddeshwar
Mallick, Tanwi
contents Existing multi-agent Large Language Model (LLM) frameworks for code generation typically use execution feedback and improve iteratively using Input/Output (I/O) test cases. However, this does not work for scientific workflows, where I/O test cases do not exist, and generating them requires solving the very problem at hand. To address this, we introduce MOSAIC, a training-free multi-agent framework for scientific code generation without I/O supervision. Instead of execution feedback, MOSAIC employs a student-teacher knowledge distillation framework that grounds generation through domain-specific examples and structured problem decomposition. To further mitigate hallucinations across chained subproblems, we introduce a Consolidated Context Window (CCW) for maintaining consistent reasoning across agents. Experiments on the SciCode benchmark show that MOSAIC improves accuracy, executability, and numerical precision over existing approaches while relying on lightweight models.
format Preprint
id arxiv_https___arxiv_org_abs_2604_23106
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle No Test Cases, No Problem: Distillation-Driven Code Generation for Scientific Workflows
Raghavan, Siddeshwar
Mallick, Tanwi
Software Engineering
Artificial Intelligence
Multiagent Systems
Existing multi-agent Large Language Model (LLM) frameworks for code generation typically use execution feedback and improve iteratively using Input/Output (I/O) test cases. However, this does not work for scientific workflows, where I/O test cases do not exist, and generating them requires solving the very problem at hand. To address this, we introduce MOSAIC, a training-free multi-agent framework for scientific code generation without I/O supervision. Instead of execution feedback, MOSAIC employs a student-teacher knowledge distillation framework that grounds generation through domain-specific examples and structured problem decomposition. To further mitigate hallucinations across chained subproblems, we introduce a Consolidated Context Window (CCW) for maintaining consistent reasoning across agents. Experiments on the SciCode benchmark show that MOSAIC improves accuracy, executability, and numerical precision over existing approaches while relying on lightweight models.
title No Test Cases, No Problem: Distillation-Driven Code Generation for Scientific Workflows
topic Software Engineering
Artificial Intelligence
Multiagent Systems
url https://arxiv.org/abs/2604.23106