@article {10.3844/jcssp.2026.2617.2632,
article_type = {journal},
title = {Empowering Small Language Models With Chain of Thought and Parameter Efficient Fine Tuning for Efficient Deep Reasoning},
author = {Raina, Aryan and Gupta, Shiwani and Jadhav, Jagruti and Bhonde, Sampada and Mathur, Shilpa and Maha, Anand and Sankhe, Pranjali},
volume = {22},
number = {8},
year = {2026},
month = {Sep},
pages = {2617-2632},
doi = {10.3844/jcssp.2026.2617.2632},
url = {https://thescipub.com/abstract/jcssp.2026.2617.2632},
abstract = {Small Language Models (SLMs) run faster and fit on modest hardware, yet solving multi-step logic problems has traditionally been difficult for them. This work investigates a systematic, multi-model framework that combines Chain-of-Thought (CoT) prompting with Low-Rank Adaptation (LoRA) parameter-efficient fine-tuning, and measures the contribution of each component independently. We evaluate three SLMs (TinyLlama-1.1B, Phi-3-mini-4k-instruct, and Qwen2.5-1.5B) using 1,164 training samples drawn from GSM8K, Microsoft Orca-Math-Word-Problems-200k, and OpenAI HumanEval, and evaluate on 500 GSM8K test problems. Phi-3-mini-4k-instruct achieves the strongest overall gains, reaching 16.0% exact match and 61.7% step accuracy under +CoT+LoRA, a 95% improvement over its own baseline. The decline observed under CoT-only prompting and the subsequent recovery under CoT+LoRA were statistically significant (McNemar test, p},
journal = {Journal of Computer Science},
publisher = {Science Publications}
}