subsumer 0.7.0__tar.gz

subsumer-0.7.0/Cargo.toml

@@ -0,0 +1,68 @@
+[package]
+name = "subsume"
+version = "0.7.0"
+edition = "2021"
+rust-version = "1.75"
+license = "MIT OR Apache-2.0"
+authors = ["Arc <attobop@gmail.com>"]
+description = "Geometric region embeddings (boxes, cones, octagons, Gaussians, hyperbolic intervals, sheaf networks) for subsumption, entailment, and logical query answering"
+keywords = ["embeddings", "box-embeddings", "geometric", "subsumption", "ontology"]
+categories = ["science", "algorithms"]
+repository = "https://github.com/arclabs561/subsume"
+documentation = "https://docs.rs/subsume"
+readme = "README.md"
+publish = true
+include = [
+    "/src/**/*.rs",
+    "/examples/**/*.rs",
+    "/docs/katex-header.html",
+    "/Cargo.toml",
+    "/LICENSE-*",
+    "/README.md",
+    "/pretrained/*.json",
+]
+
+[dependencies]
+serde = { version = "1", features = ["derive"] }
+thiserror = "2"
+rand = { version = "0.9", optional = true }
+hyperball = { version = "0.1.3", optional = true }
+skel = { version = "0.1.1", optional = true }
+ndarray = { version = "0.16", features = ["serde-1"], optional = true }
+serde_json = { version = "1.0", optional = true }
+candle-core = { version = "0.9", optional = true }
+petgraph = { version = "0.7", optional = true }
+
+[features]
+default = ["ndarray-backend"]
+ndarray-backend = ["dep:ndarray", "dep:serde_json", "rand"]
+candle-backend = ["dep:candle-core"]
+rand = ["dep:rand"]
+petgraph = ["dep:petgraph"]
+hyperbolic = ["dep:hyperball", "dep:skel", "dep:ndarray"]
+
+[dev-dependencies]
+criterion = { version = "0.5", features = ["html_reports"] }
+proptest = "1.5"
+tempfile = "3"
+
+[[example]]
+name = "hyperbolic_demo"
+required-features = ["hyperbolic"]
+
+
+[[example]]
+name = "taxobell_training"
+required-features = ["candle-backend"]
+
+[lints.rust]
+missing_docs = "warn"
+unsafe_code = "warn"
+
+[package.metadata.docs.rs]
+all-features = true
+rustdoc-args = ["--cfg", "docsrs", "--html-in-header", "docs/katex-header.html"]
+
+[lints.clippy]
+correctness = { level = "deny", priority = -1 }
+perf = { level = "warn", priority = -1 }

subsumer-0.7.0/LICENSE-APACHE

@@ -0,0 +1,203 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (which shall not include Communications that are clearly marked or
+      otherwise designated in writing by the copyright holder as "Not a Work").
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based upon (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright holder as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment format for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright 2025 arclabs561
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+   implied. See the License for the specific language governing
+   permissions and limitations under the License.
+

subsumer-0.7.0/LICENSE-MIT

@@ -0,0 +1,22 @@
+MIT License
+
+Copyright (c) 2025 arclabs561
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.
+

subsumer-0.7.0/PKG-INFO

@@ -0,0 +1,9 @@
+Metadata-Version: 2.4
+Name: subsumer
+Version: 0.7.0
+Classifier: Programming Language :: Rust
+Classifier: Programming Language :: Python :: Implementation :: CPython
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Dist: numpy>=1.20
+Summary: Geometric region embeddings for knowledge graph subsumption (Python bindings for the subsume Rust crate)
+Requires-Python: >=3.9

subsumer-0.7.0/README.md

@@ -0,0 +1,261 @@
+# subsume
+
+[![crates.io](https://img.shields.io/crates/v/subsume.svg)](https://crates.io/crates/subsume)
+[![Documentation](https://docs.rs/subsume/badge.svg)](https://docs.rs/subsume)
+[![CI](https://github.com/arclabs561/subsume/actions/workflows/ci.yml/badge.svg)](https://github.com/arclabs561/subsume/actions/workflows/ci.yml)
+
+Geometric region embeddings for subsumption, entailment, and logical query answering. Boxes, cones, octagons, Gaussians, hyperbolic intervals, and sheaf networks. Ndarray and Candle backends.
+
+![Box embedding concepts](docs/box_concepts.png)
+
+*(a) Containment: nested boxes encode taxonomic is-a relationships. (b) Gumbel soft boundary: temperature controls membership sharpness. (c) Octagon: diagonal constraints cut corners for tighter volume bounds.*
+
+## What it provides
+
+### Geometric primitives
+
+| Component | What it does |
+|---|---|
+| `Box` trait | Axis-aligned hyperrectangle: volume, containment, overlap, distance |
+| `NdarrayGumbelBox` / `CandleGumbelBox` | Probabilistic boxes via Gumbel random variables (dense gradients, no flat regions; Dasgupta et al., 2020) |
+| `NdarrayCone` | Angular cones in d-dimensional space: containment via aperture, closed under negation (Zhang & Wang, NeurIPS 2021) |
+| `NdarrayOctagon` | Axis-aligned polytopes with diagonal constraints; tighter volume bounds than boxes (Charpenay & Schockaert, IJCAI 2024) |
+| `gaussian` | Diagonal Gaussian boxes: KL divergence (asymmetric containment) and Bhattacharyya coefficient (symmetric overlap) |
+| `hyperbolic` | Poincare ball embeddings and hyperbolic box intervals (via `hyperball`; requires `hyperbolic` feature) |
+| `sheaf` | Sheaf diffusion primitives: stalks, restriction maps, Laplacian (Hansen & Ghrist 2019; Bodnar et al., ICLR 2022) |
+
+### Scoring and query answering
+
+| Component | What it does |
+|---|---|
+| `distance` | Query2Box alpha-weighted point-to-box distance (Ren et al., 2020); depth-based (RegD) and boundary distances in backend modules |
+| `fuzzy` | Fuzzy t-norms/t-conorms for logical query answering (FuzzQE, Chen et al., AAAI 2022) |
+| `el` | EL++ ontology embedding: inclusion loss, role composition, existential boxes (Box2EL/TransBox) |
+
+### Taxonomy and training
+
+| Component | What it does |
+|---|---|
+| `taxonomy` | TaxoBell-format dataset loader: `.terms`/`.taxo` parsing, train/val/test splitting |
+| `taxobell` | TaxoBell combined loss: Bhattacharyya triplet + KL containment + volume regularization + sigma clipping |
+| Training utilities | Negative sampling, AMSGrad optimizer |
+| Evaluation | MRR, Hits@k, Mean Rank, nDCG |
+
+### Backends
+
+| Component | What it does |
+|---|---|
+| `NdarrayBox` / `NdarrayGumbelBox` / `NdarrayCone` / `NdarrayOctagon` | CPU backend using `ndarray::Array1<f32>` |
+| `CandleBox` / `CandleGumbelBox` | GPU/Metal backend using `candle_core::Tensor` |
+
+The ndarray backend has full geometry support. The candle backend provides
+GPU-accelerated box operations for training workflows.
+
+## Usage
+
+```toml
+[dependencies]
+subsume = { version = "0.7", features = ["ndarray-backend"] }
+ndarray = "0.16"
+```
+
+```rust
+use subsume::ndarray_backend::NdarrayBox;
+// Renamed import avoids shadowing std::boxed::Box
+use subsume::Box as BoxRegion;
+use ndarray::array;
+
+// Box A: [0,0,0] to [1,1,1] (general concept)
+let premise = NdarrayBox::new(array![0., 0., 0.], array![1., 1., 1.], 1.0)?;
+
+// Box B: [0.2,0.2,0.2] to [0.8,0.8,0.8] (specific, inside A)
+let hypothesis = NdarrayBox::new(array![0.2, 0.2, 0.2], array![0.8, 0.8, 0.8], 1.0)?;
+
+// Containment probability: P(B inside A)
+let p = premise.containment_prob(&hypothesis)?;
+assert!(p > 0.9);
+```
+
+## Examples
+
+```bash
+cargo run -p subsume --example containment_hierarchy    # taxonomic is-a relationships with nested boxes
+cargo run -p subsume --example gumbel_box_exploration   # Gumbel boxes, soft containment, temperature effects
+cargo run -p subsume --example cone_training            # training cone embeddings on a taxonomy
+cargo run -p subsume --example box_training             # training box embeddings on a 25-entity taxonomy
+cargo run -p subsume --example taxobell_demo            # TaxoBell Gaussian box losses on a mini taxonomy
+cargo run -p subsume --example query2box                # Query2Box: multi-hop queries, box intersection, distance scoring
+cargo run -p subsume --example octagon_demo             # octagon embeddings: diagonal constraints, containment, volume
+cargo run -p subsume --example fuzzy_query              # fuzzy query answering: t-norms, De Morgan duality, rankings
+cargo run -p subsume --example dataset_training --release # full pipeline: WN18RR-format data, train, evaluate
+cargo run -p subsume --example imagenet_hierarchy --release # 252 Tiny ImageNet synsets, volume-depth correlation
+cargo run -p subsume --example save_checkpoint --release           # generate pretrained/wordnet_subset.json checkpoint
+cargo run -p subsume --features hyperbolic --example hyperbolic_demo  # Poincare ball: hierarchy preservation, exponential capacity
+cargo run -p subsume --example wn18rr_training --release  # WN18RR benchmark: 40K entities, 20 epochs
+cargo run -p subsume --example el_training              # EL++ box embeddings on a biomedical-style ontology
+cargo run -p subsume --features candle-backend --example taxobell_training  # TaxoBell MLP encoder training (Candle)
+```
+
+See [`examples/README.md`](examples/README.md) for a guide to choosing the right example.
+
+## Tests
+
+```bash
+cargo test -p subsume
+```
+
+Unit, property, and doc tests covering:
+
+- Box geometry: intersection, union, containment, overlap, distance, volume, truncation
+- Gumbel boxes: membership probability, temperature edge cases, Bessel volume
+- Cones: angular containment, negation closure, aperture bounds
+- Octagon: intersection closure, containment, Sutherland-Hodgman volume
+- Fuzzy: t-norm/t-conorm commutativity, associativity, De Morgan duality
+- Gaussian boxes, EL++ ontology losses, sheaf networks, hyperbolic geometry
+- Training: MRR, Hits@k, Mean Rank, nDCG, negative sampling, AMSGrad
+
+## Choosing a geometry
+
+| Geometry | When to use it | ¬? | Key tradeoff |
+|---|---|---|---|
+| NdarrayBox / NdarrayGumbelBox | Containment hierarchies, each dimension independent | No | Simple, fast; Gumbel adds dense gradients where hard boxes have zero gradient |
+| Cone | Multi-hop queries requiring negation (FOL with ¬) | Yes | Closed under complement; angular parameterization harder to initialize |
+| Octagon | Rule-aware KG completion; tighter containment than boxes | No | Diagonal constraints cut box corners; more parameters per entity |
+| Gaussian | Taxonomy expansion with uncertainty (TaxoBell) | No | KL = asymmetric containment; Bhattacharyya = symmetric overlap |
+| Hyperbolic | Tree-like hierarchies with exponential branching | No | Low-dim capacity; numerical care near Poincare ball boundary |
+
+## Why regions instead of points?
+
+Point embeddings (TransE, RotatE, ComplEx) represent entities as vectors. They work
+well for link prediction -- RotatE hits 0.476 MRR on WN18RR, BoxE hits 0.451.
+For standard triple scoring, points are simpler and equally accurate.
+
+Regions become necessary when the task requires structure that points cannot encode.
+The core operation is **containment probability**:
+
+$$P(B \subseteq A) = \frac{\text{Vol}(A \cap B)}{\text{Vol}(B)}$$
+
+If B fits inside A, $P = 1$. If disjoint, $P = 0$. This is the scoring
+function used for evaluation (`containment_prob`).
+
+| What you need | Points | Regions |
+|---|---|---|
+| Containment (A ⊆ B) | No -- points have no interior | Box nesting = subsumption |
+| Volume = generality | No -- points have no size | Large box = broad concept |
+| Intersection (A ∧ B) | No set operations | Box ∩ Box = another box |
+| Negation (¬A) | No complement | Cone complement = another cone |
+| Uncertainty per dimension | No | Gaussian sigma |
+
+Three tasks where point embeddings structurally fail:
+
+1. **Ontology completion (EL++)**: "Dog is-a Animal" requires representing one concept's
+   extension as a subset of another's. Points have no containment. Box2EL, TransBox, and
+   DELE use boxes for this and outperform point baselines on Gene Ontology, GALEN, and
+   Anatomy.
+
+2. **Logical query answering (∧, ∨, ¬)**: multi-hop KG queries with conjunction,
+   disjunction, and negation need set operations. ConE handles all three (MRR 52.9 on
+   FB15k EPFO+negation queries vs Query2Box's 41.0 and BetaE's 44.6). Points cannot
+   attempt negation queries at all.
+
+3. **Taxonomy expansion**: inserting a new concept at the right depth requires knowing
+   both what it is (similarity) and how general it is (volume). TaxoBell uses Gaussian
+   boxes where KL divergence gives asymmetric parent-child containment for free.
+
+If your task is link prediction or entity similarity, use RotatE. If you need
+containment, set operations, or volume, you need regions.
+
+See [`docs/SUBSUMPTION_HISTORY.md`](docs/SUBSUMPTION_HISTORY.md) for the research
+history of geometric subsumption embeddings, from hard boxes through Gumbel, cones, and beyond.
+
+## Why Gumbel boxes?
+
+![Gumbel gradient landscape](docs/gumbel_robustness.png)
+
+*(a) Membership probability at a box boundary: hard boxes have a discontinuous step, Gumbel boxes have smooth sigmoids controlled by temperature. (b) Gradient magnitude: hard boxes produce zero gradient everywhere except the exact boundary (gray regions), while Gumbel boxes provide gradients throughout the space.*
+
+Gumbel boxes model coordinates as Gumbel random variables, creating soft boundaries
+that provide dense gradients throughout training. Hard boxes create flat regions where
+gradients vanish; Gumbel boxes solve this *local identifiability problem*
+(Dasgupta et al., 2020). Lower temperature (small beta) gives crisper boundaries with
+sharper gradients; higher temperature gives broader gradients that reach further from
+the boundary but sacrifice containment precision.
+
+## Training convergence
+
+![Training convergence](docs/training_convergence.png)
+
+*25-entity taxonomy learned over 200 epochs. Left: total violation drops 3 orders of magnitude. Right: containment probabilities converge to 1.0 at different rates depending on hierarchy depth. Reproduce: `cargo run --example box_training` or `uv run scripts/plot_training.py`.*
+
+## Embedding export
+
+`BoxEmbeddingTrainer::export_embeddings()` returns flat f32 vectors suitable for
+safetensors, numpy (via reshape), and vector databases:
+
+```rust,ignore
+let (ids, mins, maxs) = trainer.export_embeddings();
+// mins/maxs are flat Vec<f32> of length n_entities * dim
+// Reshape to (n_entities, dim) for numpy/safetensors
+```
+
+Checkpoint save/load via serde:
+
+```rust,ignore
+let json = serde_json::to_string(&trainer)?;
+let restored: BoxEmbeddingTrainer = serde_json::from_str(&json)?;
+```
+
+## Integration patterns
+
+Convert from petgraph (when `petgraph` feature is enabled):
+
+```rust,ignore
+use subsume::petgraph_adapter::from_graph;
+let dataset = from_graph(&my_digraph);
+```
+
+Convert from polars (no dependency needed, user-side code):
+
+```rust,ignore
+use subsume::dataset::Triple;
+let triples: Vec<Triple> = df.column("head")?.str()?
+    .into_iter()
+    .zip(df.column("relation")?.str()?)
+    .zip(df.column("tail")?.str()?)
+    .filter_map(|((h, r), t)| Some(Triple::new(h?, r?, t?)))
+    .collect();
+let dataset = Dataset::new(triples, vec![], vec![]);
+```
+
+## References
+
+- Nickel & Kiela (2017). "Poincare Embeddings for Learning Hierarchical Representations"
+- Vilnis et al. (2018). "Probabilistic Embedding of Knowledge Graphs with Box Lattice Measures"
+- Li et al. (2019). "Smoothing the Geometry of Probabilistic Box Embeddings" (ICLR 2019)
+- Abboud et al. (2020). "BoxE: A Box Embedding Model for Knowledge Base Completion"
+- Dasgupta et al. (2020). "Improving Local Identifiability in Probabilistic Box Embeddings"
+- Ren et al. (2020). "Query2Box: Reasoning over Knowledge Graphs using Box Embeddings"
+- Hansen & Ghrist (2019). "Toward a Spectral Theory of Cellular Sheaves"
+- Bodnar et al. (2022). "Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs"
+- Boratko et al. (2021). "Box Embeddings: An open-source library for representation learning using geometric structures" (EMNLP Demo)
+- Chen et al. (2021). "Probabilistic Box Embeddings for Uncertain Knowledge Graph Reasoning" (BEUrRE, ACL 2021)
+- Gebhart, Hansen & Schrater (2021). "Knowledge Sheaves: A Sheaf-Theoretic Framework for Knowledge Graph Embedding"
+- Zhang & Wang (2021). "ConE: Cone Embeddings for Multi-Hop Reasoning over Knowledge Graphs"
+- Chen et al. (2022). "Fuzzy Logic Based Logical Query Answering on Knowledge Graphs"
+- Jackermeier et al. (2023). "Dual Box Embeddings for the Description Logic EL++"
+- Yang, Chen & Sattler (2024). "TransBox: EL++-closed Ontology Embedding"
+- Bourgaux et al. (2024). "Knowledge Base Embeddings: Semantics and Theoretical Properties" (KR 2024)
+- Charpenay & Schockaert (2024). "Capturing Knowledge Graphs and Rules with Octagon Embeddings"
+- Lacerda et al. (2024). "Strong Faithfulness for ELH Ontology Embeddings" (TGDK 2024)
+- Huang et al. (2023). "Concept2Box: Joint Geometric Embeddings for Learning Two-View Knowledge Graphs"
+- Mashkova et al. (2024). "DELE: Deductive EL++ Embeddings for Knowledge Base Completion"
+- Yang & Chen (2025). "Achieving Hyperbolic-Like Expressiveness with Arbitrary Euclidean Regions"
+- Mishra et al. (2026). "TaxoBell: Gaussian Box Embeddings for Self-Supervised Taxonomy Expansion" (WWW '26)
+
+## See also
+
+- [`hyperball`](https://crates.io/crates/hyperball) -- hyperbolic geometry primitives (optional; requires `hyperbolic` feature for Poincare ball embeddings)
+
+## License
+
+MIT OR Apache-2.0

subsumer-0.7.0/docs/katex-header.html

@@ -0,0 +1,82 @@
+<!-- KaTeX CSS -->
+<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/katex.min.css" crossorigin="anonymous">
+
+<!-- KaTeX JavaScript -->
+<!--
+Note: we intentionally do not use SRI hashes here.
+In practice these files are used in multiple contexts (rustdoc, local previews, etc.) and
+stale/incorrect integrity values fail closed and silently break rendering.
+-->
+<script defer src="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/katex.min.js" crossorigin="anonymous"></script>
+<script defer src="https://cdn.jsdelivr.net/npm/katex@0.16.9/dist/contrib/auto-render.min.js" crossorigin="anonymous"></script>
+
+<script>
+// Configure KaTeX to render math in rustdoc
+document.addEventListener("DOMContentLoaded", function() {
+    // Render math in all documentation content
+    renderMathInElement(document.body, {
+        delimiters: [
+            {left: "\\[", right: "\\]", display: true},
+            {left: "\\(", right: "\\)", display: false},
+            {left: "$$", right: "$$", display: true},
+            {left: "$", right: "$", display: false}
+        ],
+        throwOnError: false,
+        strict: false,
+        trust: true
+    });
+    
+    // Also watch for dynamically loaded content (rustdoc uses JS to load content)
+    const observer = new MutationObserver(function(mutations) {
+        mutations.forEach(function(mutation) {
+            mutation.addedNodes.forEach(function(node) {
+                if (node.nodeType === 1) { // Element node
+                    renderMathInElement(node, {
+                        delimiters: [
+                            {left: "\\[", right: "\\]", display: true},
+                            {left: "\\(", right: "\\)", display: false},
+                            {left: "$$", right: "$$", display: true},
+                            {left: "$", right: "$", display: false}
+                        ],
+                        throwOnError: false,
+                        strict: false,
+                        trust: true
+                    });
+                }
+            });
+        });
+    });
+    
+    // Observe the main content area
+    const content = document.querySelector('.content') || document.querySelector('main') || document.body;
+    if (content) {
+        observer.observe(content, {
+            childList: true,
+            subtree: true
+        });
+    }
+});
+</script>
+
+<style>
+/* Enhance KaTeX rendering in rustdoc */
+.rustdoc .katex {
+    font-size: 1.1em;
+}
+
+.rustdoc .katex-display {
+    margin: 1.5em 0;
+    overflow-x: auto;
+    overflow-y: hidden;
+}
+
+/* Better spacing for math in doc comments */
+.rustdoc .docblock .katex-display {
+    margin: 1em 0;
+}
+
+.rustdoc .docblock .katex {
+    font-size: 1.05em;
+}
+</style>
+