Trace-of-Thought Prompting: Investigating Prompt-Based Knowledge Distillation Through Question Decomposition
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866915268008083456 |
|---|---|
| author | McDonald, Tyler Emami, Ali |
| author_facet | McDonald, Tyler Emami, Ali |
| contents | Knowledge distillation allows smaller neural networks to emulate the performance of larger, teacher models with reduced computational demands. Traditional methods for Large Language Models (LLMs) often necessitate extensive fine-tuning, which limits their accessibility. To address this, we introduce Trace-of-Thought Prompting, a novel framework designed to distill critical reasoning capabilities from high-resource teacher models (over 8 billion parameters) to low-resource student models (up to 8 billion parameters). This approach leverages problem decomposition to enhance interpretability and facilitate human-in-the-loop interventions. Empirical evaluations on the GSM8K and MATH datasets show that student models achieve accuracy gains of up to 113% on GSM8K and 21% on MATH, with significant improvements particularly notable in smaller models like Llama 2 and Zephyr. Our results suggest a promising pathway for open-source, low-resource models to eventually serve both as both students and teachers, potentially reducing our reliance on high-resource, proprietary models. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_20946 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Trace-of-Thought Prompting: Investigating Prompt-Based Knowledge Distillation Through Question Decomposition McDonald, Tyler Emami, Ali Computation and Language Artificial Intelligence Knowledge distillation allows smaller neural networks to emulate the performance of larger, teacher models with reduced computational demands. Traditional methods for Large Language Models (LLMs) often necessitate extensive fine-tuning, which limits their accessibility. To address this, we introduce Trace-of-Thought Prompting, a novel framework designed to distill critical reasoning capabilities from high-resource teacher models (over 8 billion parameters) to low-resource student models (up to 8 billion parameters). This approach leverages problem decomposition to enhance interpretability and facilitate human-in-the-loop interventions. Empirical evaluations on the GSM8K and MATH datasets show that student models achieve accuracy gains of up to 113% on GSM8K and 21% on MATH, with significant improvements particularly notable in smaller models like Llama 2 and Zephyr. Our results suggest a promising pathway for open-source, low-resource models to eventually serve both as both students and teachers, potentially reducing our reliance on high-resource, proprietary models. |
| title | Trace-of-Thought Prompting: Investigating Prompt-Based Knowledge Distillation Through Question Decomposition |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2504.20946 |