ATOMIC: an Interpretable Clustering Method Based on Data Topology
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
State-of-the-art clustering algorithms are well equipped to partition a dataset, but provide no insight into their process for selecting a particular partition. As such clustering strategies which work in an interpretable manner, i.e., in a way that can be understood, are desired. Current explainable clustering approaches focus mainly on explaining k-means clustering, which limits their scope to problems where clusters are spherical and convex. Additionally, many of these algorithms struggle to produce explanations on high dimensional data due to their computational complexity. We propose an interpretable clustering methodology which addresses these challenges. Our algorithm, ATOMIC: Analysis of Topology Oriented Method for Interpretable Clustering, is designed to answer the question "why does my data fit into distinct clusters?". This is done by identifying a set of variables that are responsible for isolating a cluster of data-points from the rest in the dataset. To locate partitions which are defined by specific variables we utilize Novelty Search with Local Competition. ATOMIC relies on basic concepts from topology to partition the dataset. We test our algorithm on well-studied high-dimensional datasets, along with performance comparisons to state-of-the-art clustering methodologies. We compare our algorithm in terms of interpretability to other interpretable and explainable clustering methodologies.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it