Year: 2025
Type: preprint
Abstract: Recent progress in large language models (LLMs) has made knowledge-grounded reasoning increasingly practical, yet KG-based QA systems often pay a steep price in efficiency and transparency. In typical pipelines, symbolic paths are scored by neural encoders or repeatedly re-ranked by multiple LLM calls, which inflates latency and GPU cost and makes the decision process hard to audit. We introduce PathHD, an encoder-free framework for knowledge-graph reasoning that couples hyperdimensional computing (HDC) with a single LLM call per query. Given a query, PathHD represents relation paths as block-diagonal GHRR hypervectors, retrieves candidate paths using a calibrated blockwise cosine similarity with Top-K pruning, and then performs a one-shot LLM adjudication that outputs the final answer together with supporting, citeable paths. The design is enabled by three technical components: (i) an order-sensitive, non-commutative binding operator for composing multi-hop paths, (ii) a robust similarity calibration that stabilizes hypervector retrieval, and (iii) an adjudication stage that preserves interpretability while avoiding per-path LLM scoring. Across WebQSP, CWQ, and GrailQA, PathHD matches or improves Hits@1 compared to strong neural baselines while using only one LLM call per query, reduces end-to-end latency by $40-60\%$, and lowers GPU memory by $3-5\times$ due to encoder-free retrieval. Overall, the results suggest that carefully engineered HDC path representations can serve as an effective substrate for efficient and faithful KG-LLM reasoning, achieving a strong accuracy-efficiency-interpretability trade-off.
Source: arXiv (Cornell University)
Institution: University of California, Irvine
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
Awards: Neurally-Inspired Integration of Communication and Cognitive Computation in Hyperspace, G1875902628, CPS: Small: Advanced Hyperdimensional and Symbolic Knowledge Transfer for Cyber-Physical Systems, G2871076685, UKRI/BBSRC-NSF/BIO: Interpretable and Noise-Robust Machine Learning for Neurophysiology, G4720003262, FuSe-TG: Regaining the Edge, G5553233121, CPS: Small: Brain-Inspired Memorization and Attention for Intelligent Sensing, G6098521345, G6604890474, G8017170229, Hyperdimensional Neural Computation for Real-Time Cognitive Learning, G8876996369
OpenAlex ID: W4417290328
Cites:
Cited by:
Related to: