gradiend 0.1.0__tar.gz

gradiend-0.1.0/CITATION.cff

@@ -0,0 +1,20 @@
+cff-version: 1.2.0
+message: If you use this software, please cite both the article from preferred-citation and the software itself.
+authors:
+  - family-names: Jonathan Drechsel
+    given-names: Steffen Herbold
+title: 'GRADIEND: Feature Learning within Neural Networks Exemplified through Biases'
+version: 1.0.0
+url: https://arxiv.org/abs/2502.01406
+date-released: '2025-02-04'
+preferred-citation:
+  authors:
+    - family-names: Jonathan Drechsel
+      given-names: Steffen Herbold
+  title: 'GRADIEND: Feature Learning within Neural Networks Exemplified through Biases'
+  url: https://arxiv.org/abs/2502.01406
+  type: generic
+  year: '2025'
+  conference: {}
+  publisher: {}
+

gradiend-0.1.0/LICENSE

@@ -0,0 +1,201 @@
+                                 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
+      not 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
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (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 owner 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 syntax 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 [yyyy] [name of copyright owner]
+
+   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.

gradiend-0.1.0/MANIFEST.in

@@ -0,0 +1,6 @@
+include README.md
+include LICENSE
+include CITATION.cff
+recursive-include examples *.py *.md
+recursive-exclude * __pycache__
+recursive-exclude * *.py[co]

gradiend-0.1.0/PKG-INFO

@@ -0,0 +1,130 @@
+Metadata-Version: 2.4
+Name: gradiend
+Version: 0.1.0
+Summary: GRADIEND: Gradient-based Targeted Feature Learning within Neural Networks
+Author-email: Jonathan Drechsel <jonathan.drechsel@uni-passau.de>
+License: Apache-2.0
+Project-URL: Homepage, https://github.com/aieng-lab/gradiend
+Project-URL: Documentation, https://aieng-lab.github.io/gradiend/
+Project-URL: Repository, https://github.com/aieng-lab/gradiend
+Project-URL: Bug Tracker, https://github.com/aieng-lab/gradiend/issues
+Keywords: machine learning,neural networks,feature learning,interpretability
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Science/Research
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: torch>=2.0.0
+Requires-Dist: transformers<5.0.0,>=4.30.0
+Requires-Dist: numpy>=1.20.0
+Requires-Dist: pandas>=1.3.0
+Requires-Dist: tqdm>=4.60.0
+Provides-Extra: recommended
+Requires-Dist: safetensors>=0.3.0; extra == "recommended"
+Requires-Dist: sentencepiece>=0.1.99; extra == "recommended"
+Requires-Dist: datasets>=3.0.2; extra == "recommended"
+Requires-Dist: spacy>=3.5.0; extra == "recommended"
+Requires-Dist: pyarrow>=14.0.0; extra == "recommended"
+Requires-Dist: matplotlib>=3.5.0; extra == "recommended"
+Requires-Dist: seaborn>=0.12.0; extra == "recommended"
+Requires-Dist: venn>=0.1.0; extra == "recommended"
+Requires-Dist: matplotlib_venn>=0.11.7; extra == "recommended"
+Provides-Extra: data
+Requires-Dist: datasets>=3.0.2; extra == "data"
+Requires-Dist: spacy>=3.5.0; extra == "data"
+Requires-Dist: pyarrow>=14.0.0; extra == "data"
+Provides-Extra: plot
+Requires-Dist: matplotlib>=3.5.0; extra == "plot"
+Requires-Dist: seaborn>=0.12.0; extra == "plot"
+Requires-Dist: venn>=0.1.0; extra == "plot"
+Requires-Dist: matplotlib_venn>=0.11.7; extra == "plot"
+Provides-Extra: dev
+Requires-Dist: mkdocs-material>=9.0.0; extra == "dev"
+Requires-Dist: mkdocstrings[python]>=0.24.0; extra == "dev"
+Requires-Dist: pytest>=7.0.0; extra == "dev"
+Requires-Dist: pytest-cov>=4.0.0; extra == "dev"
+Dynamic: license-file
+
+# GRADIEND
+
+**Gradient-based targeted feature learning within neural networks.** Learn where a language model encodes a feature (e.g. gender, race) and rewrite the model to strengthen or weaken it—for example debias it—while keeping other behaviour unchanged.
+
+[![arXiv:2502.01406](https://img.shields.io/badge/arXiv-2502.01406-blue.svg)](https://arxiv.org/abs/2502.01406)
+[![arXiv:2601.09313](https://img.shields.io/badge/arXiv-2601.09313-blue.svg)](https://arxiv.org/abs/2601.09313)
+[![Python 3.9+](https://img.shields.io/badge/python-3.9+-blue.svg)](https://www.python.org/downloads/)
+[![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
+[![Tests](https://github.com/aieng-lab/gradiend/actions/workflows/tests.yml/badge.svg)](https://github.com/aieng-lab/gradiend/actions/workflows/tests.yml)
+[![Documentation](https://img.shields.io/badge/docs-github.io-blue.svg)](https://aieng-lab.github.io/gradiend/)
+
+Paper: [GRADIEND: Feature Learning within Neural Networks Exemplified through Biases](https://arxiv.org/abs/2502.01406) 
+
+---
+
+## Install
+
+```bash
+pip install gradiend
+```
+
+Optional extras (install one or combine with e.g. `gradiend[data,plot]`):
+
+| Tag | Command | Includes                                     |
+|-----|---------|----------------------------------------------|
+| **recommended** | `pip install gradiend[recommended]` | full data generation, training, and plotting |
+| **data** | `pip install gradiend[data]` | for advanced data generation and import |
+| **plot** | `pip install gradiend[plot]` | for plotting support |                                                                                   |
+| **dev** | `pip install gradiend[dev]` | for building docs and running tests |
+
+---
+
+## Links
+
+| | |
+|---|---|
+| **Documentation** | [aieng-lab.github.io/gradiend](https://aieng-lab.github.io/gradiend/) |
+| **Source & issues** | [github.com/aieng-lab/gradiend](https://github.com/aieng-lab/gradiend) |
+| **Example scripts & notebooks** | [gradiend/examples](https://github.com/aieng-lab/gradiend/tree/main/gradiend/examples) — start_workflow, english_pronouns, gender_de, gender_en, race_religion, etc. |
+| **Datasets (Hugging Face)** | [de-gender-case-articles](https://huggingface.co/datasets/aieng-lab/de-gender-case-articles), [gradiend_race_data](https://huggingface.co/datasets/aieng-lab/gradiend_race_data), [gradiend_religion_data](https://huggingface.co/datasets/aieng-lab/gradiend_religion_data), [biasneutral](https://huggingface.co/datasets/aieng-lab/biasneutral), [geneutral](https://huggingface.co/datasets/aieng-lab/geneutral) |
+| **Pre-trained GRADIEND models** | [bert-base-cased-gradiend-gender-debiased](https://huggingface.co/aieng-lab/bert-base-cased-gradiend-gender-debiased), [gpt2-gradiend-gender-debiased](https://huggingface.co/aieng-lab/gpt2-gradiend-gender-debiased), [Llama-3.2-3B-gradiend-gender-debiased](https://huggingface.co/aieng-lab/Llama-3.2-3B-gradiend-gender-debiased) |
+
+---
+
+## Citation
+
+```bibtex
+@misc{drechsel2025gradiend,
+  title={{GRADIEND}: Feature Learning within Neural Networks Exemplified through Biases},
+  author={Jonathan Drechsel and Steffen Herbold},
+  year={2025},
+  eprint={2502.01406},
+  archivePrefix={arXiv},
+  primaryClass={cs.LG},
+  url={https://arxiv.org/abs/2502.01406},
+}
+```
+
+German definite articles study:
+
+```bibtex
+@misc{drechsel2026understanding,
+  title={Understanding or Memorizing? A Case Study of German Definite Articles in Language Models},
+  author={Jonathan Drechsel and Erisa Bytyqi and Steffen Herbold},
+  year={2026},
+  eprint={2601.09313},
+  archivePrefix={arXiv},
+  primaryClass={cs.CL},
+  url={https://arxiv.org/abs/2601.09313},
+}
+```
+
+---
+
+## License
+
+Apache 2.0. See [LICENSE](https://github.com/aieng-lab/gradiend/blob/main/LICENSE) on GitHub.

gradiend-0.1.0/PYPI.md

@@ -0,0 +1,77 @@
+# GRADIEND
+
+**Gradient-based targeted feature learning within neural networks.** Learn where a language model encodes a feature (e.g. gender, race) and rewrite the model to strengthen or weaken it—for example debias it—while keeping other behaviour unchanged.
+
+[![arXiv:2502.01406](https://img.shields.io/badge/arXiv-2502.01406-blue.svg)](https://arxiv.org/abs/2502.01406)
+[![arXiv:2601.09313](https://img.shields.io/badge/arXiv-2601.09313-blue.svg)](https://arxiv.org/abs/2601.09313)
+[![Python 3.9+](https://img.shields.io/badge/python-3.9+-blue.svg)](https://www.python.org/downloads/)
+[![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
+[![Tests](https://github.com/aieng-lab/gradiend/actions/workflows/tests.yml/badge.svg)](https://github.com/aieng-lab/gradiend/actions/workflows/tests.yml)
+[![Documentation](https://img.shields.io/badge/docs-github.io-blue.svg)](https://aieng-lab.github.io/gradiend/)
+
+Paper: [GRADIEND: Feature Learning within Neural Networks Exemplified through Biases](https://arxiv.org/abs/2502.01406) 
+
+---
+
+## Install
+
+```bash
+pip install gradiend
+```
+
+Optional extras (install one or combine with e.g. `gradiend[data,plot]`):
+
+| Tag | Command | Includes                                     |
+|-----|---------|----------------------------------------------|
+| **recommended** | `pip install gradiend[recommended]` | full data generation, training, and plotting |
+| **data** | `pip install gradiend[data]` | for advanced data generation and import |
+| **plot** | `pip install gradiend[plot]` | for plotting support |                                                                                   |
+| **dev** | `pip install gradiend[dev]` | for building docs and running tests |
+
+---
+
+## Links
+
+| | |
+|---|---|
+| **Documentation** | [aieng-lab.github.io/gradiend](https://aieng-lab.github.io/gradiend/) |
+| **Source & issues** | [github.com/aieng-lab/gradiend](https://github.com/aieng-lab/gradiend) |
+| **Example scripts & notebooks** | [gradiend/examples](https://github.com/aieng-lab/gradiend/tree/main/gradiend/examples) — start_workflow, english_pronouns, gender_de, gender_en, race_religion, etc. |
+| **Datasets (Hugging Face)** | [de-gender-case-articles](https://huggingface.co/datasets/aieng-lab/de-gender-case-articles), [gradiend_race_data](https://huggingface.co/datasets/aieng-lab/gradiend_race_data), [gradiend_religion_data](https://huggingface.co/datasets/aieng-lab/gradiend_religion_data), [biasneutral](https://huggingface.co/datasets/aieng-lab/biasneutral), [geneutral](https://huggingface.co/datasets/aieng-lab/geneutral) |
+| **Pre-trained GRADIEND models** | [bert-base-cased-gradiend-gender-debiased](https://huggingface.co/aieng-lab/bert-base-cased-gradiend-gender-debiased), [gpt2-gradiend-gender-debiased](https://huggingface.co/aieng-lab/gpt2-gradiend-gender-debiased), [Llama-3.2-3B-gradiend-gender-debiased](https://huggingface.co/aieng-lab/Llama-3.2-3B-gradiend-gender-debiased) |
+
+---
+
+## Citation
+
+```bibtex
+@misc{drechsel2025gradiend,
+  title={{GRADIEND}: Feature Learning within Neural Networks Exemplified through Biases},
+  author={Jonathan Drechsel and Steffen Herbold},
+  year={2025},
+  eprint={2502.01406},
+  archivePrefix={arXiv},
+  primaryClass={cs.LG},
+  url={https://arxiv.org/abs/2502.01406},
+}
+```
+
+German definite articles study:
+
+```bibtex
+@misc{drechsel2026understanding,
+  title={Understanding or Memorizing? A Case Study of German Definite Articles in Language Models},
+  author={Jonathan Drechsel and Erisa Bytyqi and Steffen Herbold},
+  year={2026},
+  eprint={2601.09313},
+  archivePrefix={arXiv},
+  primaryClass={cs.CL},
+  url={https://arxiv.org/abs/2601.09313},
+}
+```
+
+---
+
+## License
+
+Apache 2.0. See [LICENSE](https://github.com/aieng-lab/gradiend/blob/main/LICENSE) on GitHub.

gradiend-0.1.0/README.md

@@ -0,0 +1,227 @@
+# GRADIEND
+
+[![arXiv:2502.01406](https://img.shields.io/badge/arXiv-2502.01406-blue.svg)](https://arxiv.org/abs/2502.01406)
+[![arXiv:2601.09313](https://img.shields.io/badge/arXiv-2601.09313-blue.svg)](https://arxiv.org/abs/2601.09313)
+[![Python 3.8+](https://img.shields.io/badge/python-3.8+-blue.svg)](https://www.python.org/downloads/)
+[![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
+[![Tests](https://github.com/aieng-lab/gradiend/actions/workflows/tests.yml/badge.svg)](https://github.com/aieng-lab/gradiend/actions/workflows/tests.yml)
+[![Documentation](https://img.shields.io/badge/docs-github.io-blue.svg)](https://aieng-lab.github.io/gradiend/)
+
+GRADIEND (Gradient-based targeted feature learning within neural networks) learns features inside language models by training an encoder-decoder on gradients. Find where a model encodes a feature (e.g. gender, race) and rewrite the model to strengthen or weaken it—for example debias it—while keeping other behaviour. See [GRADIEND: Feature Learning within Neural Networks Exemplified through Biases](https://arxiv.org/abs/2502.01406).
+
+![GRADIEND overview](docs/img/workflow-diagram.png)
+
+
+## Installation
+
+With minimal requirements (enables full data generation, training and evaluation workflow):
+```bash
+pip install gradiend
+```
+
+With recommended dependencies (enables additional features such as Hugging Face datasets, safetensors, and plotting):
+
+```bash
+pip install gradiend[recommended]
+```
+
+Similar, `gradiend[data]` and `gradiend[plot]` only adds additional dependencies for data generation and plotting, respectively.
+
+Installation from source:
+
+```bash
+git clone https://github.com/aieng-lab/gradiend.git
+cd gradiend
+pip install -e ".[recommended]"
+```
+
+## Quick start
+
+Self-contained example (runs with the few sentences below; for real use, add more base data):
+
+```python
+from gradiend import TextPredictionDataCreator, TextFilterConfig, TrainingArguments, TextPredictionTrainer
+
+base = [
+    "The chef tasted the soup, then he added pepper.",
+    "The players ran; they scored.",
+    "She handed the package to the courier and asked them to deliver it.",
+    # ... see more texts in the dropdown below (needed to run a meaningful training)
+]
+creator = TextPredictionDataCreator(
+    base_data=base,
+    feature_targets=[
+        TextFilterConfig(targets=["he", "she", "it"], id="3SG"),
+        TextFilterConfig(target="they", id="3PL"),
+    ],
+)
+training = creator.generate_training_data(max_size_per_class=30)
+neutral = creator.generate_neutral_data(additional_excluded_words=["i", "we", "you"], max_size=10)
+
+args = TrainingArguments(train_batch_size=2, eval_steps=5, max_steps=20, learning_rate=1e-3)
+trainer = TextPredictionTrainer(model="bert-base-cased", data=training, eval_neutral_data=neutral, args=args)
+trainer.train()
+trainer.plot_training_convergence()
+
+enc_result = trainer.evaluate_encoder(plot=True)
+dec = trainer.evaluate_decoder(plot=True, target_class="3SG")
+changed_model = trainer.rewrite_base_model(decoder_results=dec, target_class="3SG")
+```
+
+<details>
+<summary><strong>More base texts required to run above example </strong> </summary>
+
+```python
+base = [
+    "The chef tasted the soup, then he added a pinch of pepper and stirred it.",
+    "The pianist closed her eyes and played the final chord; she had practised it for weeks.",
+    "The dog ran to the door; it wanted to go outside and chase the ball.",
+    "The report will be ready for review by the end of the week.",
+    "His phone rang twice before he picked it up and left the room.",
+    "She handed the package to the courier and asked them to deliver it by noon.",
+    "The players on both teams huddled on the pitch before they ran back to their positions.",
+    "The author read a paragraph from his novel to the audience and then they asked him questions.",
+    "Breakfast is included for all guests staying at the hotel.",
+    "The nurse checked her clipboard and told the family members they could all visit soon.",
+    "The birds gathered on the wire; when the cat moved they flew away at once.",
+    "The mechanic wiped his hands and said the car would be ready by Friday; he had fixed it.",
+    "She opened the window and watched the leaves fall in the garden below.",
+    "The committee members met on Tuesday and they voted to postpone the decision.",
+    "When the referee blew the whistle he showed the player a yellow card.",
+    "The volunteers packed the boxes and said they would all load the van at dawn.",
+    "Her brother and his colleagues sent a text saying they were stuck in traffic together.",
+    "The dog barked at the postman and he dropped the letters when it lunged.",
+    "The meeting ran over time so the chair cut the last two items; the participants agreed they would meet again.",
+    "The weather improved by the afternoon and the streets dried quickly.",
+    "Rain poured all morning but the basement stayed dry; the pump had run all night and it held.",
+    "She found the recipe in the drawer; it calls for butter, flour, and a pinch of salt.",
+    "The road was closed for repairs and the local officials said they expected the detour would add twenty minutes.",
+    "The film ended at midnight and the group of friends all went home; they shared umbrellas in the rain.",
+    "The museum opens at nine and closes at six on weekdays.",
+    "Trains run every fifteen minutes during peak hours.",
+    "The recipe works best when the oven is preheated properly.",
+    "The meeting has been moved to the smaller room on the third floor.",
+    "The gardener pruned the roses and he left the cuttings by the gate.",
+    "The drummer lost her stick in the middle of the song but she kept the beat.",
+    "One of the students raised her hand and asked them to repeat the question.",
+    "The board members announced that they had approved the budget; the CEO signed it.",
+    "He left the book on the table and she noticed the door was open when it swung.",
+    "The car broke down on the motorway and it had to be towed away.",
+    "The cat jumped off the wall; it landed in the flower bed and ran off.",
+    "Two colleagues from his team offered him a seat but he preferred to stand; she nodded and they sat down.",
+    "Coffee was served in the lobby while the conference continued upstairs.",
+    "Keys were left on the counter by the front door.",
+    "Parking is available in the lot behind the building.",
+    "The driver signalled left and he turned into the car park.",
+    "She replied to the email before the other team members had a chance to follow up, and they thanked her later.",
+    "The laptop was slow so it needed a restart and more memory.",
+    "The lecture will be recorded and posted online by tomorrow.",
+    "His keys fell under the seat and he had to reach for them.",
+    "The team members celebrated after they won the final match.",
+    "The manager gave his approval and then the project team met; they scheduled the launch.",
+    "They invited her to the meeting and she accepted on the spot.",
+    "Lunch will be served in the canteen from twelve to two.",
+    "The doctor checked her notes and told the patients they could go home; they thanked her at the door.",
+    "The fans in the crowd cheered when they saw the result on the screen.",
+    "When the alarm went off he switched it off and got up.",
+    "The staff members finished the inventory and they reported the count.",
+    "Her colleague forwarded the file and the two of them opened it together; they checked every page.",
+    "The cat stretched and it jumped onto the sofa.",
+    "The council members met last night and they approved the new bylaws.",
+    "The coach gave his feedback and the players practised the drill again; they repeated it three times.",
+    "The intern made her first presentation and her colleagues in the room asked a few questions; they praised her work.",
+    "The dog waited by the bowl; it had not been fed yet.",
+    "The panel members discussed the proposal and they reached a consensus.",
+    "He locked the office and she set the alarm before they left the building together.",
+    "The van pulled up and it unloaded the delivery at the back.",
+    "The results are published on the intranet every Friday.",
+    "The neighbours across the street waved to her and she waved back as they all passed by.",
+    "Snow fell all day but the gritters had been out and it cleared.",
+    "The schedule is on the wall next to the break room.",
+    "The bus was late so her friends and she missed the start of the film; they had to sneak into their seats.",
+    "The document is in the shared folder and can be edited by anyone.",
+    "The waiter brought the bill and he left the tip on the table.",
+    "The singer forgot the words but she carried on and the audience members applauded; they gave her a standing ovation.",
+    "The printer ran out of paper and it stopped mid-job.",
+    "The deadline has been extended to the end of the month.",
+    "The committee will reconvene next week to finalise the report.",
+    "Tea and biscuits are provided in the kitchen on each floor.",
+    "She booked the room and the hotel staff confirmed the reservation by email; they also sent her directions.",
+    "The gate was left open so the horse got out and it wandered off.",
+    "The contract is valid for twelve months from the signing date.",
+]
+```
+</details>
+
+<p align="center">
+  <img src="docs/img/readme_training_convergence.png"
+       width="32%"
+       style="vertical-align: middle;" />
+  <img src="docs/img/readme_encoder_analysis_split_test.png"
+       width="32%"
+       style="vertical-align: middle;" />
+  <img src="docs/img/readme_decoder_probability_shifts_3SG.png"
+       width="32%"
+       style="vertical-align: middle;" />
+</p>
+
+
+
+More examples: [gradiend/examples](https://github.com/aieng-lab/gradiend/tree/main/gradiend/examples)
+
+## Documentation
+
+- [Documentation](https://aieng-lab.github.io/gradiend/)
+- [Installation details](docs/installation.md)
+- [Train Your first GRADIEND Model](docs/start.md) — minimal workflow
+- [Tutorials](docs/index.md#tutorials)
+- [API reference](docs/api-reference.md)
+
+## Examples
+
+Example scripts and notebooks: [gradiend/examples](https://github.com/aieng-lab/gradiend/tree/main/gradiend/examples) on GitHub (not in the pip package; download a file or read to get inspired).
+
+- [start_workflow.py](gradiend/examples/start_workflow.py) — Minimal runnable example
+- [english_pronouns.ipynb](gradiend/examples/english_pronouns.ipynb) — English pronouns (3SG vs 3PL): data creation from Wikipedia → training → evaluation ([script](gradiend/examples/english_pronouns.py))
+- [gender_de.py](gradiend/examples/gender_de.py) — German gender (masc_nom vs fem_nom)
+- [gender_en.py](gradiend/examples/gender_en.py) — English gender with name augmentation
+- [gender_de_decoder_only.py](gradiend/examples/gender_de_decoder_only.py) — Decoder-only model with optional MLM head
+- [race_religion.py](gradiend/examples/race_religion.py) — Race and religion bias
+
+## Datasets and models
+
+**Datasets (Hugging Face):** [de-gender-case-articles](https://huggingface.co/datasets/aieng-lab/de-gender-case-articles), [gradiend_race_data](https://huggingface.co/datasets/aieng-lab/gradiend_race_data), [gradiend_religion_data](https://huggingface.co/datasets/aieng-lab/gradiend_religion_data), [biasneutral](https://huggingface.co/datasets/aieng-lab/biasneutral), [geneutral](https://huggingface.co/datasets/aieng-lab/geneutral), and more.
+
+**Pre-trained GRADIEND models:** [bert-base-cased-gradiend-gender-debiased](https://huggingface.co/aieng-lab/bert-base-cased-gradiend-gender-debiased), [gpt2-gradiend-gender-debiased](https://huggingface.co/aieng-lab/gpt2-gradiend-gender-debiased), [Llama-3.2-3B-gradiend-gender-debiased](https://huggingface.co/aieng-lab/Llama-3.2-3B-gradiend-gender-debiased), and others.
+
+## Citation
+
+```bibtex
+@misc{drechsel2025gradiend,
+  title={{GRADIEND}: Feature Learning within Neural Networks Exemplified through Biases},
+  author={Jonathan Drechsel and Steffen Herbold},
+  year={2025},
+  eprint={2502.01406},
+  archivePrefix={arXiv},
+  primaryClass={cs.LG},
+  url={https://arxiv.org/abs/2502.01406},
+}
+```
+
+For the German definite articles study using GRADIEND:
+
+```bibtex
+@misc{drechsel2026understanding,
+  title={Understanding or Memorizing? A Case Study of German Definite Articles in Language Models},
+  author={Jonathan Drechsel and Erisa Bytyqi and Steffen Herbold},
+  year={2026},
+  eprint={2601.09313},
+  archivePrefix={arXiv},
+  primaryClass={cs.CL},
+  url={https://arxiv.org/abs/2601.09313},
+}
+```
+
+## License
+
+Apache 2.0. See [LICENSE](LICENSE).