HAPPI: Hyperbolic Hierarchical Part Prototypes for Image Recognition
Bibliographic record
Abstract
Prototypical part networks have gained prominence in com-puter vision due to their inherent interpretability, enabling decisions based on representative part features without post-hoc explanations. However, existing prototypical net-works learn part-based features in flat Euclidean space, yet they could better capture the natural hierarchical relationships within image features to enhance performance on tasks requiring structural understanding. To address this opportunity, we propose HAPPI (Hierarchical And Part Prototypical Image recognition), a framework that lever-ages hyperbolic geometry to organize prototypical part fea-tures hierarchically within a Lorentzian manifold. By ar-ranging localized generic features near the hyperboloid origin and broader specific features farther away, HAPPI learns generic prototypes for defining local patterns and specific prototypes that aggregate broader discriminative cues, effectively capturing hierarchy in image data. Our approach is model-agnostic and can be applied to various prototypical neural networks and backbones. We evaluate HAPPI on several baselines and datasets, showing that hy-perbolic prototypes match or outperform Euclidean ones while adding interpretability. Qualitative results reveal that generic prototypes highlight localized, class-defining traits, while specific prototypes capture broader patterns across larger regions, enabling differentiation through both lo-cal and contextual features. Our code can be found at http://github.com/DeepRCL/HAPPI.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.004 | 0.002 |
| Research integrity | 0.002 | 0.002 |
| Insufficient payload (model declined to judge) | 0.008 | 0.004 |
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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".