# A Geometric View of SRC: Learning Representations for Stable Residual Inference

Source: [A Geometric View of SRC: Learning Representations for Stable Residual Inference](<https://www.sota2.com/research/paper/6a191dcfdbd7150506f82a61>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-30T05:00:11.986Z

Authors: Vangelis P. Oikonomou

Published: 2026-05-28T09:33:18.000Z

## Abstract

Reconstruction-based inference assigns a class by comparing class-wise reconstruction residuals; Sparse Representation Classification (SRC) is a canonical instance whose reliability depends on the geometry of the learned representation. We adopt a strict training-inference separation: SRC is used only as a fixed test-time rule and is never differentiated, unrolled, or optimized during training. In a span-level idealization based on class-conditional spans and their associated projection residuals, we formalize residual-ordering stability through a residual margin and characterize geometric obstructions -- span overlap, dominance, and near-overlap via small principal angles -- that can collapse this margin in worst-case directions. This span-level theory is primary: it specifies when the idealized residual family is well-separated, and it provides a conditional solver-level interpretation for practical residual approximations (e.g., OMP) insofar as they remain close to the span-level residual ordering. Under explicit coverage and separation assumptions, we derive a quantitative lower bound on the (idealized) residual margin. Guided by these targets, we propose geometry-shaping objectives that promote masked within-class self-expressiveness, discourage cross-class reconstruction pathways and inter-class span alignment, and prevent collapse -- without invoking SRC residuals or predictions during training. Experiments on images (COIL-100), text (TREC), and EEG connectivity evaluate all representations under identical fixed SRC/OMP inference and report residual margins and geometric diagnostics; cross-entropy is included only as a reference geometry under the same evaluation protocol.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [AD vs HC classification on EEG AD vs HC (summary)](<https://www.sota2.com/research/sota/ad-vs-hc-classification-on-eeg-ad-vs-hc-summary.md>): AD vs. HC classification; EEG AD vs HC (summary); SRC Balanced Accuracy: 87.1.

- [Image Classification on COIL-100 (SRC Acc, Margin, Rank, Cohesion)](<https://www.sota2.com/research/sota/image-classification-on-coil-100-src-acc-margin-rank-cohesion.md>): Image Classification; COIL-100; SRC Accuracy: 100.

- [Question Classification on TREC (SRC Acc, Top-2 Margin, Eff. Rank, Cohesionmax)](<https://www.sota2.com/research/sota/question-classification-on-trec-src-acc-top-2-margin-eff-rank-cohesionmax.md>): Question Classification; TREC; SRC Accuracy: 94.6.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2605.29673>)

[PDF](<https://arxiv.org/pdf/2605.29673.pdf>)
