Bridging the Reasoning Gap in Vietnamese with Small Language Models via Test-Time Scaling

TL;DR

This paper explores test-time scaling and fine-tuning to improve reasoning in Vietnamese Small Language Models, demonstrating significant performance gains and analyzing trade-offs in prompting strategies.

Contribution

It introduces Vi-S1K and Vi-Elementary-Bench datasets, and shows that supervised fine-tuning combined with test-time scaling enhances reasoning in resource-constrained models.

Findings

Supervised fine-tuning improves explanation quality by 77%.

Base model has high latent knowledge but formatting issues.

Structured prompting can degrade performance due to cognitive load.

Abstract

The democratization of ubiquitous AI hinges on deploying sophisticated reasoning capabilities on resource-constrained devices. However, Small Language Models (SLMs) often face a "reasoning gap", particularly in non-English languages like Vietnamese, where they struggle to maintain coherent chains of thought. This paper investigates Test-Time Scaling strategies for the Qwen3-1.7B architecture within the context of Vietnamese Elementary Mathematics. We introduce Vi-S1K, a high-fidelity reasoning dataset localized via a Gemini 2.5 Flash-Lite powered pipeline, and Vi-Elementary-Bench, a dual-resource benchmark for rigorous evaluation. Using an LLM-as-a-Judge protocol, we reveal that the base model possesses robust latent knowledge (Accuracy: 4.05/5.00) but suffers from a severe "formatting gap" in communication. Supervised Fine-Tuning (SFT) acts as a critical "reasoning unlocker", yielding a 77% improvement in Explanation Quality and bridging the gap between raw calculation and pedagogical coherence. Furthermore, our analysis of prompting strategies uncovers a significant trade-off: structured frameworks like ReAct impose a "cognitive tax" on the 1.7B parameter capacity, degrading performance relative to pure Chain-of-Thought (CoT) combined with Self-Consistency. These findings establish a deployment hierarchy for SLMs, demonstrating that SFT combined with simplified test-time scaling is superior to complex agentic workflows for edge-based reasoning.