Enhancing Small Language Models Reasoning through Knowledge Graph Grounding

Researchers enhanced small language models (SLMs) like Gemma 3 and Llama 3.2 by grounding them in knowledge graphs using a neuro-symbolic agentic framework. The approach improved reasoning on the CLUTRR benchmark by 1.5-2x but faced challenges from extraction errors and a "distraction effect" from noisy self-generated facts.

arXiv:2607.14149v1 Announce Type: new Abstract: Although large language models LLMs have set benchmarks for zero-shot reasoning, their deployment remains cost-prohibitive and environmentally taxing. Small Language Models SLMs offer a sustainable alternative, but prone to errors, on tasks requiring complex, multi-hop logical grounding. We investigate a neuro-symbolic agentic framework to enhance the reasoning capabilities of SLMs, specifically Gemma 3 1B, 4B and Llama 3.2 3B , using the CLUTRR kinship benchmark. Our approach transforms the SLM into a minimalist agent utilizing two specialized tool calls: extract facts for symbolic triplet extraction and get hint for expert reasoning via a Relational Graph Convolutional Network RGCN . We evaluate these models across two configurations, both in an Oracle scenario with ground-truth triplets and a Realistic scenario relying on self-extracted knowledge. Our results reveal that while RGCN-derived hints provide a 1.5 - 2x performance gain over story-only baselines, the system is constrained by the extraction bottleneck and sequential deductive fragility, where early extraction errors compound over multi-hop chains. Furthermore, we identify a "distraction effect" in specific architectures where noisy, self-generated facts degrade performance despite the presence of expert hints. This work characterizes the challenges of symbolic grounding in low-resource agentic systems and provides a roadmap for iterative verification in neuro-symbolic agentic pipelines.