[PRE REVIEW]: LangFair: A Python Package for Assessing Bias and Fairness in Large Language Model Use Cases

[editorialbot]

Submitting author: @dylanbouchard (Dylan Bouchard)
Repository: https://github.com/cvs-health/langfair
Branch with paper.md (empty if default branch): joss_paper
Version: v0.2.0
Editor: @crvernon
Reviewers: @xavieryao, @emily-sexton
Managing EiC: Chris Vernon

Status

Status badge code:

HTML: <a href="https://joss.theoj.org/papers/0fbc09de7fa4e873ac68c3cb30afdd66"><img src="https://joss.theoj.org/papers/0fbc09de7fa4e873ac68c3cb30afdd66/status.svg"></a>
Markdown: [![status](https://joss.theoj.org/papers/0fbc09de7fa4e873ac68c3cb30afdd66/status.svg)](https://joss.theoj.org/papers/0fbc09de7fa4e873ac68c3cb30afdd66)

Author instructions

Thanks for submitting your paper to JOSS @dylanbouchard. Currently, there isn't a JOSS editor assigned to your paper.

@dylanbouchard if you have any suggestions for potential reviewers then please mention them here in this thread (without tagging them with an @). You can search the list of people that have already agreed to review and may be suitable for this submission.

Editor instructions

The JOSS submission bot @editorialbot is here to help you find and assign reviewers and start the main review. To find out what @editorialbot can do for you type:

@editorialbot commands

[editorialbot]

Hello human, I'm @editorialbot, a robot that can help you with some common editorial tasks.

For a list of things I can do to help you, just type:

@editorialbot commands

For example, to regenerate the paper pdf after making changes in the paper's md or bib files, type:

@editorialbot generate pdf

[editorialbot]

Software report:

github.com/AlDanial/cloc v 1.90  T=0.47 s (201.6 files/s, 241798.1 lines/s)
-------------------------------------------------------------------------------
Language                     files          blank        comment           code
-------------------------------------------------------------------------------
JSON                            11              0              0          99452
Python                          58            695           1920           3399
Jupyter Notebook                11              0           6138            834
TeX                              2            112            100            587
Markdown                         7            161              0            419
TOML                             1              5              0             57
YAML                             3              9             10             52
CSV                              2              0              0              2
-------------------------------------------------------------------------------
SUM:                            95            982           8168         104802
-------------------------------------------------------------------------------

Commit count by author:

    70	Dylan Bouchard
     6	Viren Bajaj
     5	zeya30
     4	Mohit Singh Chauhan
     2	David Skarbrevik
     2	Zeya Ahmad
     2	virenbajaj
     1	Vasistha

[editorialbot]

Paper file info:

📄 Wordcount for paper.md is 2267

✅ The paper includes a Statement of need section

[editorialbot]

License info:

🟡 License found: Other (Check here for OSI approval)

[editorialbot]

Reference check summary (note 'MISSING' DOIs are suggestions that need verification):

✅ OK DOIs

- 10.48550/ARXIV.2205.09209 is OK
- 10.1145/3442188.3445924 is OK
- 10.18653/v1/2021.naacl-main.191 is OK
- 10.1016/j.simpa.2024.100619 is OK
- 10.5281/zenodo.12608602 is OK
- 10.1016/j.metrad.2023.100017 is OK
- 10.18653/v1/N19-3002 is OK
- 10.1145/3351095.3372851 is OK
- 10.24963/ijcai.2019/199 is OK
- 10.1609/aaai.v37i12.26691 is OK
- 10.3115/1073083.1073135 is OK
- 10.1145/3604915.3608860 is OK
- 10.1109/ICDM.2012.45 is OK
- 10.18653/v1/2022.naacl-main.122 is OK

🟡 SKIP DOIs

- No DOI given, and none found for title: Gender Bias in Coreference Resolution
- No DOI given, and none found for title: GitHub - vnmssnhv/NeuTralRewriter: Neutral rewrite...
- No DOI given, and none found for title: Mind the GAP: A Balanced Corpus of Gendered Ambigu...
- No DOI given, and none found for title: Unmasking Contextual Stereotypes: Measuring and Mi...
- No DOI given, and none found for title: GitHub - katyfelkner/winoqueer
- No DOI given, and none found for title: GitHub - umanlp/RedditBias: Code & Data for the pa...
- No DOI given, and none found for title: Saif | Bias EEC
- No DOI given, and none found for title: GitHub - HowieHwong/TrustGPT: Can we Trust Large L...
- No DOI given, and none found for title: GitHub - nyu-mll/BBQ: Repository for the Bias Benc...
- No DOI given, and none found for title: Holistic Evaluation of Language Models
- No DOI given, and none found for title: DecodingTrust: A Comprehensive Assessment of Trust...
- No DOI given, and none found for title: GitHub - huggingface/evaluate: Evaluate: A library...
- No DOI given, and none found for title: Beyond the Imitation Game: Quantifying and extrapo...
- No DOI given, and none found for title: TrustLLM: Trustworthiness in Large Language Models
- No DOI given, and none found for title: LightEval: A lightweight framework for LLM evaluat...
- No DOI given, and none found for title: GitHub - BCG-X-Official/artkit: Automated prompt-b...
- No DOI given, and none found for title: GitHub - confident-ai/deepeval: The LLM Evaluation...
- No DOI given, and none found for title: GitHub - Giskard-AI/giskard: Open-Source Evaluatio...
- No DOI given, and none found for title: AI Fairness 360:  An Extensible Toolkit for Detect...
- No DOI given, and none found for title: Fairlearn: Assessing and Improving Fairness of AI ...
- No DOI given, and none found for title: Aequitas: A Bias and Fairness Audit Toolkit
- No DOI given, and none found for title: GitHub - tensorflow/fairness-indicators: Tensorflo...
- No DOI given, and none found for title: LiFT: A Scalable Framework for Measuring Fairness ...
- No DOI given, and none found for title: The LinkedIn Fairness Toolkit (LiFT)
- No DOI given, and none found for title: An Actionable Framework for Assessing Bias and Fai...
- No DOI given, and none found for title: Large Language Models: A Survey
- No DOI given, and none found for title: Towards Auditing Large Language Models: Improving ...
- No DOI given, and none found for title: Counterfactual Fairness
- No DOI given, and none found for title: Bias and Fairness in Large Language Models: A Surv...
- No DOI given, and none found for title: ROUGE: A Package for Automatic Evaluation of Summa...
- No DOI given, and none found for title: Equality of Opportunity in Supervised Learning
- No DOI given, and none found for title: A Reductions Approach to Fair Classification

❌ MISSING DOIs

- 10.18653/v1/n18-2003 may be a valid DOI for title: Gender Bias in Coreference Resolution: Evaluation ...
- 10.18653/v1/2021.findings-emnlp.211 may be a valid DOI for title: Collecting a Large-Scale Gender Bias Dataset for C...
- 10.18653/v1/2021.acl-long.416 may be a valid DOI for title: StereoSet: Measuring stereotypical bias in pretrai...
- 10.18653/v1/2020.emnlp-main.154 may be a valid DOI for title: CrowS-Pairs: A Challenge Dataset for Measuring Soc...
- 10.1609/aaai.v34i05.6267 may be a valid DOI for title: On Measuring and Mitigating Biased Inferences of W...
- 10.18653/v1/2020.findings-emnlp.301 may be a valid DOI for title: RealToxicityPrompts: Evaluating Neural Toxic Degen...
- 10.18653/v1/2022.ltedi-1.4 may be a valid DOI for title: Measuring Harmful Sentence Completion in Language ...
- 10.18653/v1/2020.findings-emnlp.311 may be a valid DOI for title: UnQovering Stereotyping Biases via Underspecified ...
- 10.1145/3576840.3578295 may be a valid DOI for title: Grep-BiasIR: a dataset for investigating gender re...
- 10.1109/tvcg.2019.2934619 may be a valid DOI for title: The What-If Tool: Interactive Probing of Machine L...
- 10.18653/v1/2020.findings-emnlp.7 may be a valid DOI for title: Reducing Sentiment Bias in Language Models via Cou...
- 10.1145/2783258.2783311 may be a valid DOI for title: Certifying and removing disparate impact
- 10.1007/978-3-642-33486-3_3 may be a valid DOI for title: Fairness-aware classifier with prejudice remover r...
- 10.1145/3603195.3603198 may be a valid DOI for title: On Fairness and Calibration
- 10.1145/3278721.3278779 may be a valid DOI for title: Mitigating Unwanted Biases with Adversarial Learni...
- 10.18653/v1/2021.acl-long.150 may be a valid DOI for title: Intrinsic Bias Metrics Do Not Correlate with Appli...

❌ INVALID DOIs

- https://doi.org/10.1016/j.iotcps.2023.04.003 is INVALID because of 'https://doi.org/' prefix

[editorialbot]

👉📄 Download article proof 📄 View article proof on GitHub 📄 👈

[editorialbot]

Five most similar historical JOSS papers:

Jury: A Comprehensive Evaluation Toolkit
Submitting author: @devrimcavusoglu
Handling editor: @crvernon (Active)
Reviewers: @evamaxfield, @KennethEnevoldsen
Similarity score: 0.5992

FAT Forensics: A Python Toolbox for Implementing and
Deploying Fairness, Accountability and Transparency
Algorithms in Predictive Systems
Submitting author: @So-Cool
Handling editor: @arokem (Retired)
Reviewers: @bernease, @osolari
Similarity score: 0.5970

matbench-genmetrics: A Python library for benchmarking crystal structure generative models using time-based splits of Materials Project structures
Submitting author: @sgbaird
Handling editor: @phibeck (Active)
Reviewers: @ml-evs, @mkhorton, @jamesrhester
Similarity score: 0.5933

Contextualized: Heterogeneous Modeling Toolbox
Submitting author: @cnellington
Handling editor: @fabian-s (Active)
Reviewers: @holl-, @pescap
Similarity score: 0.5818

TorchGAN: A Flexible Framework for GAN Training and Evaluation
Submitting author: @avik-pal
Handling editor: @arfon (Active)
Reviewers: @NMontanaBrown, @terrytangyuan
Similarity score: 0.5785

⚠️ Note to editors: If these papers look like they might be a good match, click through to the review issue for that paper and invite one or more of the authors before considering asking the reviewers of these papers to review again for JOSS.

[crvernon]

@editorialbot assign me as editor

👋 @dylanbouchard - I'll take this one on myself. While I am getting you some reviewers, please take care of the following:

• Reduce the paper length to around 1000 words
• Please add in the missing DOI listed above in your bib file where applicable

Thanks!

[editorialbot]

Assigned! @crvernon is now the editor

[dylanbouchard]

@editorialbot assign me as editor

👋 @dylanbouchard - I'll take this one on myself. While I am getting you some reviewers, please take care of the following:

• Reduce the paper length to around 1000 words
• Please add in the missing DOI listed above in your bib file where applicable

Thanks!

Hi @crvernon! I have reduced the word count by ~800 words and fixed the DOI in bib file where applicable. Please let me know if reducing the paper length further is required. Thank you very much!

[dylanbouchard]

Submitting author: @dylanbouchard (Dylan Bouchard) Repository: https://github.com/cvs-health/langfair Branch with paper.md (empty if default branch): joss_paper Version: v0.2.0 Editor: @crvernon Reviewers: Pending Managing EiC: Chris Vernon

Status

Status badge code:

HTML: <a href="https://joss.theoj.org/papers/0fbc09de7fa4e873ac68c3cb30afdd66"><img src="https://joss.theoj.org/papers/0fbc09de7fa4e873ac68c3cb30afdd66/status.svg"></a>
Markdown: [![status](https://joss.theoj.org/papers/0fbc09de7fa4e873ac68c3cb30afdd66/status.svg)](https://joss.theoj.org/papers/0fbc09de7fa4e873ac68c3cb30afdd66)

Author instructions

Thanks for submitting your paper to JOSS @dylanbouchard. Currently, there isn't a JOSS editor assigned to your paper.

@dylanbouchard if you have any suggestions for potential reviewers then please mention them here in this thread (without tagging them with an @). You can search the list of people that have already agreed to review and may be suitable for this submission.

Editor instructions

The JOSS submission bot @editorialbot is here to help you find and assign reviewers and start the main review. To find out what @editorialbot can do for you type:

@editorialbot commands

For reviewers, it seems that the following individuals may be a good fit:

• Erik-Jan van Kesteren
• Guanqun Yang
• Madiha Mirza

[crvernon]

👋 @guanqun-yang - Would you be willing to review this submission to JOSS? We carry out our checklist-driven reviews here in GitHub issues and follow these guidelines: https://joss.readthedocs.io/en/latest/review_criteria.html

[crvernon]

👋 @xavieryao - Would you be willing to review this submission to JOSS? We carry out our checklist-driven reviews here in GitHub issues and follow these guidelines: https://joss.readthedocs.io/en/latest/review_criteria.html

[crvernon]

👋 @emily-sexton - Would you be willing to review this submission to JOSS? We carry out our checklist-driven reviews here in GitHub issues and follow these guidelines: https://joss.readthedocs.io/en/latest/review_criteria.html

[xavieryao]

👋 @xavieryao - Would you be willing to review this submission to JOSS? We carry out our checklist-driven reviews here in GitHub issues and follow these guidelines: https://joss.readthedocs.io/en/latest/review_criteria.html

Happy to help. I will start reviewing right away.

[crvernon]

@editorialbot add @xavieryao as reviewer

Let me secure one more reviewer and I'll push this over to a full review thread. That will be where you can create your checklist.

[editorialbot]

@xavieryao added to the reviewers list!

[crvernon]

@editorialbot add @emily-sexton as reviewer

[editorialbot]

@emily-sexton added to the reviewers list!

[crvernon]

@editorialbot start review

👋 - Alright @dylanbouchard, @xavieryao, and @emily-sexton - I am going to close this Pre-Review and kick off the full review which you should receive a notification for. Thanks!

[editorialbot]

OK, I've started the review over in #7570.