Research

Paper

AI LLM February 24, 2026

Prompt-Level Distillation: A Non-Parametric Alternative to Model Fine-Tuning for Efficient Reasoning

Authors

Sanket Badhe, Deep Shah

Abstract

Advanced reasoning typically requires Chain-of-Thought prompting, which is accurate but incurs prohibitive latency and substantial test-time inference costs. The standard alternative, fine-tuning smaller models, often sacrifices interpretability while introducing significant resource and operational overhead. To address these limitations, we introduce Prompt-Level Distillation (PLD). We extract explicit reasoning patterns from a Teacher model and organize them into a structured list of expressive instructions for the Student model's System Prompt. Evaluated on the StereoSet and Contract-NLI datasets using Gemma-3 4B, PLD improved Macro F1 scores from 57\% to 90.0\% and 67\% to 83\% respectively, enabling this compact model to match frontier performance with negligible latency overhead. These expressive instructions render the decision-making process transparent, allowing for full human verification of logic, making this approach ideal for regulated industries such as law, finance, and content moderation, as well as high-volume use cases and edge devices.

Metadata

arXiv ID: 2602.21103
Provider: ARXIV
Primary Category: cs.CL
Published: 2026-02-24
Fetched: 2026-02-25 06:05

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2602.21103v1</id>\n    <title>Prompt-Level Distillation: A Non-Parametric Alternative to Model Fine-Tuning for Efficient Reasoning</title>\n    <updated>2026-02-24T17:03:21Z</updated>\n    <link href='https://arxiv.org/abs/2602.21103v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2602.21103v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Advanced reasoning typically requires Chain-of-Thought prompting, which is accurate but incurs prohibitive latency and substantial test-time inference costs. The standard alternative, fine-tuning smaller models, often sacrifices interpretability while introducing significant resource and operational overhead. To address these limitations, we introduce Prompt-Level Distillation (PLD). We extract explicit reasoning patterns from a Teacher model and organize them into a structured list of expressive instructions for the Student model's System Prompt. Evaluated on the StereoSet and Contract-NLI datasets using Gemma-3 4B, PLD improved Macro F1 scores from 57\\% to 90.0\\% and 67\\% to 83\\% respectively, enabling this compact model to match frontier performance with negligible latency overhead. These expressive instructions render the decision-making process transparent, allowing for full human verification of logic, making this approach ideal for regulated industries such as law, finance, and content moderation, as well as high-volume use cases and edge devices.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CL'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.IR'/>\n    <published>2026-02-24T17:03:21Z</published>\n    <arxiv:primary_category term='cs.CL'/>\n    <author>\n      <name>Sanket Badhe</name>\n    </author>\n    <author>\n      <name>Deep Shah</name>\n    </author>\n  </entry>"
}