Arxiv 2509 25045 Hyperdimensional Probe Decoding Llm Representations Via Vect
ASecurity
We introduce the Hyperdimensional Probe, a hybrid supervised probe that combines symbolic representations with neural probing. Leveraging Vector Symbolic Architectures (VSAs) and hypervector algebra, it unifies prior methods: the top-down interpretability of supervised probes, SAE's sparsity-driven proxy space, and output-oriented logit investigation. Our approach enables deeper input-focused feature extraction while supporting output-oriented analysis across different LLMs, embedding sizes, and
Installs into .claude/skills of the current project.
Are you the author of Arxiv 2509 25045 Hyperdimensional Probe Decoding Llm Representations Via Vect?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/hiyenwong-arxiv-2509-25045-hyperdimensional-probe-decoding-l)
---
title: "Hyperdimensional Probe: Decoding LLM Representations via Vector Symbolic Architectures"
authors: "Marco Bronzini, Carlo Nicolini, Bruno Lepri, Jacopo Staiano, Andrea Passerini"
arxiv_id: "2509.25045"
categories: "cs.CL; cs.AI; cs.LG"
utility: 0.92
date_added: "2026-09-29"
category: "general-ml"
---
# Hyperdimensional Probe: Decoding LLM Representations via Vector Symbolic Architectures
## Abstract
We introduce the Hyperdimensional Probe, a hybrid supervised probe that combines symbolic representations with neural probing. Leveraging Vector Symbolic Architectures (VSAs) and hypervector algebra, it unifies prior methods: the top-down interpretability of supervised probes, SAE's sparsity-driven proxy space, and output-oriented logit investigation. Our approach enables deeper input-focused feature extraction while supporting output-oriented analysis across different LLMs, embedding sizes, and configurations.
## Key Contributions
- Novel approach in general ml domain
- Utility score: 0.92
- Published on arXiv: 2509.25045
## Potential Applications
- Research reference for general ml
- Building block for related systems
