Update README.md

README.md

@@ -57,7 +57,7 @@ The perturber is a seq2seq controlled generation model that rewrites text along
 The perturber takes in (i) a source text snippet, (ii) a word in the snippet referring to a demographic group, and (iii) a new target demographic attribute, and generates a perturbed snippet that refers to the target demographic attribute, while preserving overall meaning.
 
 - **Repository:** https://github.com/facebookresearch/ResponsibleNLP/
-- **Paper:** https://arxiv.org/abs/2205.12586
+- **Paper:** https://aclanthology.org/2022.emnlp-main.646/
 - **Point of Contact:** rebeccaqian@meta.com, ccross@meta.com, douwe@huggingface.co, adinawilliams@meta.com
 - **License:** MIT
 
@@ -77,47 +77,72 @@ Below we show some example inputs and outputs for the perturber rewriting text a
 Model inputs follow the format `[selected_word][target_attribute] <PERT_SEP> [input_text]`, where `selected_word` is a word that contains demographic information, `target_attribute` is a demographic attribute such as "man" or "asian", and `input_text` is the text sequence to rewrite.
 
 Currently the perturber supports text rewriting along three axes and several attributes:
-- **gender:** man, woman, non-binary
-- **race:** black, white, asian, hispanic, native-american, pacific-islander
-- **age:** child, young, middle-aged, senior, adult
+- **gender:** `man`, `woman`, `non-binary`
+- **race:** `black`, `white`, `asian`, `hispanic`, `native-american`, `pacific-islander`
+- **age:** `child`, `young`, `middle-aged`, `senior`, `adult`
  
 ### Gender
 _Input:_
-his, woman <PERT_SEP> Jack was passionate about rock climbing and his love for the sport was infectious to all men around him.
+`his, woman <PERT_SEP> Jack was passionate about rock climbing and his love for the sport was infectious to all men around him.`
 
 _Output:_
-Jackie was passionate about rock climbing and her love for the sport was infectious to all men around her.  
+`Jackie was passionate about rock climbing and her love for the sport was infectious to all men around her.`
+<br/>
 <br/>
-
 _Input:_
-Alice, man <PERT_SEP> To her girlfriend Jen, Alice was a doting mother, loving girlfriend and talented actress.
+`Alice, man <PERT_SEP> To her girlfriend Jen, Alice was a doting mother, loving girlfriend and talented actress.`
 
 _Output:_
-To his girlfriend Jen, Alan was a doting father, loving partner and talented actor.
+`To his girlfriend Jen, Alan was a doting father, loving partner and talented actor.`
 <br/>
 <br/>
-
 _Input:_
-his, non-binary <PERT_SEP> Jack was passionate about rock climbing and his love for the sport was infectious to all men around him.
+`his, non-binary <PERT_SEP> Jack was passionate about rock climbing and his love for the sport was infectious to all men around him.`
 
 _Output:_
-Jack was passionate about rock climbing and their love for the sport was infectious to all men around them.
+`Jack was passionate about rock climbing and their love for the sport was infectious to all men around them.`
 <br/>
 <br/>
 ### Age
 _Input:_
-child, senior <PERT_SEP> The young child is naive and his innocence must be protected at all costs.
+`child, senior <PERT_SEP> The young child is naive and his innocence must be protected at all costs.`
 
 _Output:_
-The elderly person is naive and his innocence must be protected at all costs.
+`The elderly person is naive and his innocence must be protected at all costs.`
 
 ### Race/Ethnicity
 _Input:_
-Asian, black <PERT_SEP> The Asian students association often hosted anime nights and boba events on campus.
+`Asian, black <PERT_SEP> The Asian students association often hosted anime nights and boba events on campus.`
 
 _Output:_
-The Black students association often hosted anime nights and boba events on campus.
-
-### Contributions
+`The Black students association often hosted anime nights and boba events on campus.`
+
+## Bias, Risks & Limitations
+Limitations of the perturber include inherent biases in demographic categorization, data sourcing and crowdsourced data collection, and the ambiguous nature of fairness and perturbability. Ambiguous instances include names, where annotators may have different preconceptions about whether they contain ethnic information. Our crowdworkers and researchers are primarily English speaking and US-based, which may introduce additional cultural biases.
+
+For an in-depth discussion of bias, risks and limitations, see the Limitations section of [our paper](https://aclanthology.org/2022.emnlp-main.646/).
+
+## Citation
+```
+@inproceedings{qian-etal-2022-perturbation,
+    title = "Perturbation Augmentation for Fairer {NLP}",
+    author = "Qian, Rebecca  and
+      Ross, Candace  and
+      Fernandes, Jude  and
+      Smith, Eric Michael  and
+      Kiela, Douwe  and
+      Williams, Adina",
+    booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing",
+    month = dec,
+    year = "2022",
+    address = "Abu Dhabi, United Arab Emirates",
+    publisher = "Association for Computational Linguistics",
+    url = "https://aclanthology.org/2022.emnlp-main.646",
+    pages = "9496--9521",
+    abstract = "Unwanted and often harmful social biases are becoming ever more salient in NLP research, affecting both models and datasets. In this work, we ask whether training on demographically perturbed data leads to fairer language models. We collect a large dataset of human annotated text perturbations and train a neural perturbation model, which we show outperforms heuristic alternatives. We find that (i) language models (LMs) pre-trained on demographically perturbed corpora are typically more fair, and (ii) LMs finetuned on perturbed GLUE datasets exhibit less demographic bias on downstream tasks, and (iii) fairness improvements do not come at the expense of performance on downstream tasks. Lastly, we discuss outstanding questions about how best to evaluate the (un)fairness of large language models. We hope that this exploration of neural demographic perturbation will help drive more improvement towards fairer NLP.",
+}
+```
+
+### Model Card Contact
 
 Thanks to [@Rebecca-Qian](https://github.com/Rebecca-Qian) for adding this model.