Zero-Direction Probing: A Linear-Algebraic Framework for Deep Analysis of Large-Language-Model Drift

Fuente: arXiv
Saved in:
Bibliographic Details
Main Author: Pandey, Amit
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912529962237952
author Pandey, Amit
author_facet Pandey, Amit
contents We present Zero-Direction Probing (ZDP), a theory-only framework for detecting model drift from null directions of transformer activations without task labels or output evaluations. Under assumptions A1--A6, we prove: (i) the Variance--Leak Theorem, (ii) Fisher Null-Conservation, (iii) a Rank--Leak bound for low-rank updates, and (iv) a logarithmic-regret guarantee for online null-space trackers. We derive a Spectral Null-Leakage (SNL) metric with non-asymptotic tail bounds and a concentration inequality, yielding a-priori thresholds for drift under a Gaussian null model. These results show that monitoring right/left null spaces of layer activations and their Fisher geometry provides concrete, testable guarantees on representational change.
format Preprint
id arxiv_https___arxiv_org_abs_2508_06776
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Zero-Direction Probing: A Linear-Algebraic Framework for Deep Analysis of Large-Language-Model Drift
Pandey, Amit
Machine Learning
Artificial Intelligence
We present Zero-Direction Probing (ZDP), a theory-only framework for detecting model drift from null directions of transformer activations without task labels or output evaluations. Under assumptions A1--A6, we prove: (i) the Variance--Leak Theorem, (ii) Fisher Null-Conservation, (iii) a Rank--Leak bound for low-rank updates, and (iv) a logarithmic-regret guarantee for online null-space trackers. We derive a Spectral Null-Leakage (SNL) metric with non-asymptotic tail bounds and a concentration inequality, yielding a-priori thresholds for drift under a Gaussian null model. These results show that monitoring right/left null spaces of layer activations and their Fisher geometry provides concrete, testable guarantees on representational change.
title Zero-Direction Probing: A Linear-Algebraic Framework for Deep Analysis of Large-Language-Model Drift
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2508.06776