Liesenfeld, Andreas, Alianda Lopez, and Mark Dingemanse. 2023. “Opening up ChatGPT: Tracking Openness, Transparency, and Accountability in Instruction-Tuned Text Generators.” In Proceedings of the 5th International Conference on Conversational User Interfaces. Eindhoven. doi:10.1145/3571884.3604316. (PDF)
Large language models that exhibit instruction-following behaviour represent one of the biggest recent upheavals in conversational interfaces, a trend in large part fuelled by the release of OpenAI's ChatGPT, a proprietary large language model for text generation fine-tuned through reinforcement learning from human feedback (LLM+RLHF). We review the risks of relying on proprietary software and survey the first crop of open-source projects of comparable architecture and functionality. The main contribution of this paper is to show that openness is differentiated, and to offer scientific documentation of degrees of openness in this fast-moving field. We evaluate projects in terms of openness of code, training data, model weights, reinforcement learning data, licensing, scientific documentation, and access methods. We find that while there is a fast-growing list of projects billing themselves as 'open source', many inherit undocumented data of dubious legality, few share the all-important RLHF components (a key site where human labour is involved), and careful scientific documentation is exceedingly rare. Degrees of openness are relevant to fairness and accountability at all points, from data collection and curation to model architecture, and from training and fine-tuning to release and deployment.
We classify projects for their degrees of openness across a predefined set of criteria in the areas of Availability, Documentation and Access. The criteria are described in detail here.
Availability | Documentation | Access |
---|---|---|
|
|
|
If you find any of this useful, please cite our work:
@inproceedings{liesenfeld_dingemanse_2024,
author = {Liesenfeld, Andreas and Dingemanse, Mark},
title = {Rethinking open source generative AI: open washing and the EU AI Act},
year = {2024},
isbn = {9798400704505},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3630106.3659005},
doi = {10.1145/3630106.3659005},
pages = {1774–1787},
numpages = {14},
keywords = {Technology assessment, large language models, text generators, text-to-image generators},
location = {, Rio de Janeiro, Brazil, },
series = {FAccT '24}
}
@inproceedings{liesenfeld_opening_2023,
address = {Eindhoven},
title = {Opening up {ChatGPT}: tracking openness, transparency, and accountability in instruction-tuned text generators},
url = {https://opening-up-chatgpt.github.io},
doi = {10.1145/3571884.3604316},
booktitle = {Proceedings of the 5th {International} {Conference} on {Conversational} {User} {Interfaces}},
publisher = {Association for Computing Machinery},
author = {Liesenfeld, Andreas and Lopez, Alianda and Dingemanse, Mark},
year = {2023},
pages = {1--6},
}
If you know of a new instruction-tuned LLM+RLHF model we should be including, you can also add an issue.
How to contribute to the live table:
yaml
file in /projectsThe live table is updated whenever there is a change to the files in the /projects/ folder.
We try to be fairly systematic in our coverage of LLM+RLHF models, documenting degrees of openness for >10 features. There are many other resources that provide more free-form listings of relevant stuff or that offer ways to interact with (open) LLMs:
Here are some background readings on why openness matters, why closed models make bad baselines, and why some of us call for more counterfoil research in times of hype:
Contributions welcome! Read the contribution guidelines first.
List of contributors:
Made with contrib.rocks.