KG-Hopper enables 7B-parameter models to outperform 70B systems on complex Knowledge Graph reasoning by embedding the entire multi-hop process into a single 'thinking' stage.
March 24, 2026
Original Paper
KG-Hopper: Empowering Compact Open LLMs with Knowledge Graph Reasoning via Reinforcement Learning
arXiv · 2603.21440
The Takeaway
It moves beyond fragile step-by-step pipelines to a unified RL-trained reasoning process. This allows compact, open-source models to match proprietary giants (GPT-4o) on knowledge-intensive tasks while remaining data-efficient.
From the abstract
Large Language Models (LLMs) demonstrate impressive natural language capabilities but often struggle with knowledge-intensive reasoning tasks. Knowledge Base Question Answering (KBQA), which leverages structured Knowledge Graphs (KGs) exemplifies this challenge due to the need for accurate multi-hop reasoning. Existing approaches typically perform sequential reasoning steps guided by predefined pipelines, restricting flexibility and causing error cascades due to isolated reasoning at each step.