mlsanityData to 14 Oct 2026

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

arXiv 2604.23106 ↗First posted 25 Apr 2026cs.SE cs.AI cs.MAcited by 0

On the map: LLM agents and tool use → Software engineering coding agents

Abstract (arXiv)

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.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use. Builds on 15 of its 22 references that are in the map. First posted on arXiv 25 Apr 2026.

Builds on

15 of 22 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation16 Aug

Other references · 14

  1. 2015Distilling the Knowledge in a Neural Networkcited by 10,324
  2. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  3. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,276
  4. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  5. 2021Program Synthesis with Large Language Modelscited by 3,637
  6. 2021Measuring Coding Challenge Competence With APPScited by 902
  7. 2022CodeGen: An Open Large Language Model for Code with Multi-Turn Program Synthesiscited by 885
  8. 2023StarCoder: may the source be with you!cited by 799
  9. 2023The Impact of Large Language Models on Scientific Discovery: a Preliminary Study using GPT-4cited by 112
  10. 2024A Comprehensive Survey of Scientific Large Language Models and Their Applications in Scientific Discoverycited by 96
  11. 2024LLM Hallucinations in Practical Code Generation: Phenomena, Mechanism, and Mitigationcited by 93
  12. 2023LLMs and the Abstraction and Reasoning Corpus: Successes, Failures, and the Importance of Object-based Representationscited by 82
2 more
  1. 2024LLMs Will Always Hallucinate, and We Need to Live With Thiscited by 78
  2. 2023MathChat: Converse to Tackle Challenging Math Problems with LLM Agentscited by 68

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025MOSAIC: Multi-agent Orchestration for Task-Intelligent Scientific Coding
  2. 2025Guided Code Generation with LLMs: A Multi-Agent Framework for Complex Code Tasks
  3. 2024MapCoder: Multi-Agent Code Generation for Competitive Problem Solving
  4. 2026CURATE: Leveraging LLM Agents to Compose, Catalog, and Deploy Reproducible Workflows
  5. 2025CodeCoR: An LLM-Based Self-Reflective Multi-Agent Framework for Code Generation
  6. 2026SiliconMind-V1: Multi-Agent Distillation and Debug-Reasoning Workflows for Verilog Code Generation
  7. 2025SEW: Self-Evolving Agentic Workflows for Automated Code Generation
  8. 2025RECODE-H: A Benchmark for Research Code Development with Interactive Human Feedback