Spaces:
Runtime error
Runtime error
Nathan Butters
commited on
Update Readme.md
Browse filesAdded the core information about the project for anyone who decides they want to be involved.
README.md
CHANGED
@@ -1,2 +1,36 @@
|
|
1 |
-
#
|
2 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Evaluating LLMs on Hugging Face
|
2 |
+
The AVID (AI Vulnerability Database) team is examining a few large language models (LLMs) on Hugging Face. We will develop a way to evaluate and catalog their vulnerabilities in the hopes of encouraging the community to contribute. As a first step, we’re going to pick a single model and try to evaluate it for vulnerabilities on a specific task. Once we have done one model, we’ll see if we can generalize our data sets and tools to function broadly on the Hugging Face platform.
|
3 |
+
|
4 |
+
## Vision
|
5 |
+
Build a foundation for evaluating LLMs using the Hugging Face platform and start populating our database with real incidents.
|
6 |
+
|
7 |
+
## Goals
|
8 |
+
* Build, test, and refine our own data sets for evaluating models
|
9 |
+
* Identify existing data sets we want to use for evaluating models (Ex. Stereoset, wino_bias, etc.)
|
10 |
+
* Test different tools and methods for evaluating LLMs so we can start to create and support some for cataloging vulnerabilities in our database
|
11 |
+
* Start populating the database with known, verified, and discovered vulnerabilities for models hosted on Hugging Face
|
12 |
+
|
13 |
+
## Resources
|
14 |
+
The links below should help anyone who wants to support the project find a place to start. They are not exhaustive, and people should feel free to add anything relevant.
|
15 |
+
* [Huggingface.co](https://huggingface.co/) - platform for hosting data sets, models, etc.
|
16 |
+
* [Papers With Code](https://paperswithcode.com/) - a platform for the ML community to share research, it may have additional data sets or papers
|
17 |
+
* Potential Models
|
18 |
+
* [xlm-roberta-base](https://huggingface.co/xlm-roberta-base)
|
19 |
+
* [Bert-base-uncased](https://huggingface.co/bert-base-uncased)
|
20 |
+
* [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased)
|
21 |
+
* [gpt2](https://huggingface.co/gpt2)
|
22 |
+
* Data Sets
|
23 |
+
* [StereoSet](https://stereoset.mit.edu/) - StereoSet is a dataset that measures stereotype bias in language models. StereoSet consists of 17,000 sentences that measure model preferences across gender, race, religion, and profession.
|
24 |
+
* [Wino_bias](https://huggingface.co/datasets/wino_bias) - WinoBias, a Winograd-schema dataset for coreference resolution focused on gender bias.
|
25 |
+
* [Jigsaw_unintended_bias](https://huggingface.co/distilbert-base-uncased) - The main target for this dataset is toxicity prediction. Several toxicity subtypes are also available, so the dataset can be used for multi-attribute prediction.
|
26 |
+
* [BigScienceBiasEval/bias-shades](https://huggingface.co/datasets/BigScienceBiasEval/bias-shades) - This dataset was curated by hand-crafting stereotype sentences by native speakers from the culture which is being targeted. (Seems incomplete)
|
27 |
+
* [md_gender_bias](https://huggingface.co/datasets/md_gender_bias) - The dataset can be used to train a model for classification of various kinds of gender bias.
|
28 |
+
* [social_bias_frames](https://huggingface.co/datasets/social_bias_frames) - This dataset supports both classification and generation. Sap et al. developed several models using the SBIC.
|
29 |
+
* [BIG-bench/keywords_to_tasks.md at main](https://github.com/google/BIG-bench/blob/main/bigbench/benchmark_tasks/keywords_to_tasks.md#pro-social-behavior) - includes many options for testing bias of different types (gender, religion, etc.)
|
30 |
+
* [FB Fairscore](https://github.com/facebookresearch/ResponsibleNLP/tree/main/fairscore) - Has a wide selection of sources, focuses on gender (including non-binary).
|
31 |
+
* Papers
|
32 |
+
* [Evaluate & Evaluation on the Hub: Better Best Practices for Data and Model Measurement](https://arxiv.org/abs/2210.01970)
|
33 |
+
* [On the Dangers of Stochastic Parrots](https://dl.acm.org/doi/10.1145/3442188.3445922)
|
34 |
+
* [Language (Technology) is Power: A Critical Survey of “Bias” in NLP](https://aclanthology.org/2020.acl-main.485/)
|
35 |
+
* [Measuring Fairness with Biased Rulers: A Comparative Study on Bias Metrics for Pre-trained Language Models](https://aclanthology.org/2022.naacl-main.122/)
|
36 |
+
* [Harms of Gender Exclusivity and Challenges in Non-Binary Representation in Language Technologies](https://aclanthology.org/2021.emnlp-main.150/)
|