Topological Attribution Distance (TAD): Revealing Segment-Level RAG Influence on LLM Output Geometry for Incident Log Analysis

2026-08-17

AI Summary

This study addresses the critical challenge of trustworthy provenance and evidence attribution for LLM-generated content in cybersecurity applications. To this end, we propose Topological Attribution Distance (TAD), a novel segment-level attribution mechanism grounded in topological geometry. By leveraging embedding space modeling and hidden state analysis, TAD quantifies the global geometric influence of retrieval logs on generated outputs. This approach adaptively localizes critical logs to enable segment-wise evidence verification and interpretable decision tracing. Consequently, TAD effectively elucidates the intrinsic mechanisms of Retrieval-Augmented Generation, providing both theoretical foundations and technical pathways for enhancing the trustworthiness of model outputs in security-critical domains.

Abstract

Large Language Models (LLMs) are increasingly being deployed in cybersecurity operations to assist cybersecurity analysts with rapid decision-making against emerging threats. However, there is a main criteria that must be met when using LLMs in cybersecurity, that is, trust in the generated outputs. As Agentic AI is integrated into operational systems, a robust evidence attribution and provenance tracking technique is essential to trace the origins of model generations. When autonomous agents make a decision (right or wrong), the ability to trace back through the decision chain is critical, as without it, teams cannot identify which segment of the data caused the model generation. Existing methods often struggle to distinguish among complex and highly similar evidence sources, such as cyber incident logs. This reveals a key gap: current approaches do not adequately capture the holistic geometric relationship between the retrieved evidence and the generated response for reliable evidence verification. To bridge this gap, we propose Topological Attribution Distance (TAD), inspired by Topology, to characterize and capture the global geometric shape of an output and its changes against its retrieved logs. In other words, if the embeddings of a specific source log drastically changes the geometry of the model's response in the embedding space, this suggests that such log is a critical source for the model's generated response. Therefore, TAD is powered by segment-level ablation attribution to investigate incident logs of an actual cyberattack. We demonstrate how TAD finds the most attributed logs on LLM outputs in an adaptive manner. This can provide an explainable and trustworthy tracing based on each LLM's hidden state to understand how geometrically different retrieved logs influence the model generation, and provide evidence verification in cybersecurity and Agentic-AI workflows.

Problem

• Evidence Attribution
• Provenance Tracking
• LLM Trustworthiness
• Cybersecurity Incident Logs
• Explainability

Innovation

• Topological Attribution Distance
• Segment-Level Ablation Attribution
• Output Geometry
• Evidence Provenance Tracking
• Cybersecurity Incident Logs