I am SO EXCITED about a FEW different ethics-relevant things on Hugging Face. BIG THING AS OF *TODAY*: GATING of datasets/models with *manual approval* infrastructure, now supported. 😍 This is HUGE for DATA GOVERNANCE in Open ML communities huggingface.co/docs/hub/model…

v excited to finally announce our new work that formalizes one of the most effective practices for training LLMs—something that many industry leaders have conspicuously avoided discussing

the paper announcements after the #emnlp anonymity deadline feels like there was a virtual conference with thousands of accepted papers! #NLProc

AAAI announces a new way to release code and trained models associated with accepted papers. Through the use of an AI Pubs RAIL License from the RAIL Initiative, authors can elect to release their code and trained models under terms that permit free and open access but are

Responsible AI Pubs Licenses are built on a model for behavioral-use that aims to reduce the risk of negative outcomes and misuse of AI. These licenses are similar to licenses used by BigScience and over 6000+ related repositories for AI models and code. Community feedback is

Developers, researchers, & policymakers have been brainstorming ways to establish responsible #AI practices — is Responsible AI Licensing (RAIL) a solution? Join us on July 18 for a panel discussion on the benefits, limitations, & opportunities of #RAIL: eventbrite.com/e/responsible-…

Open Responsible AI Licenses (Open RAIL) are licenses designed to permit free and open access, re-use, and downstream distribution of derivatives of AI artifacts as long as the behavioral-use restrictions always apply. Please consider this as a new way to share your work with the

This has been a really exciting week for #RAIL (licenses.ai) - First #Llama2 was released with a RAIL-like license with behavioral usage restrictions and now the AllenNLP (AI2) sets out their position on adopting OpenRAIL-DMs (lnkd.in/grwJ6r_3).

Given the waves of lawsuits filed against AI firms, our member Jenny Lee wrote a blog post providing some legal context for AI developers. Check it out! licenses.ai/blog/2023/8/11…

We are so excited to host a panel of speakers for our upcoming #cscw2023 workshop on data licensing, web scraping, and creative commons. If you are attending CSCW this year, add our workshop to your conference registration to receive our Zoom link!

Summer Internship: Interested in working on methods for reducing LLM hallucinations in conversations, developing suitable prompt-controls for enterprise-grade chat solutions, enabling trusted responses in LLMs and more? DM me! lnkd.in/gPN3-Sug lnkd.in/gfXEJQ-5

Exciting times: we are working on the next generation of StarCoder trained on a new dataset! 🚀 If you would like to have your code excluded from the training run you can check if your data is in the dataset and follow the link to opt-out: huggingface.co/spaces/bigcode…

Sunehri Masjid, near Udyog Bhavan There are many Sunehri masjids in Delhi, but its the one near Udyog bhavan, from the Mughal era, that is under threat of demolition, to improve traffic situation. When New Delhi was being planned many old buildings were demolished by the

Are you building and evaluating RAG systems? Presenting InspectorRAGet arxiv.org/abs/2404.17347 a platform for easily analyzing overall performance, instance level analysis, comprehensive metrics, and multiple models and more!

The RAIL License Generator (licenses.ai/rail-license-g…) is now live. If you have an #AI/#ML project consider using it to generate a license with customized use clauses to prevent harms.

See our ICML'24 position paper for more information: arxiv.org/abs/2402.05979

🌟New Benchmark! 🌟 Do you work on RAG? Are you interested in Multi-Turn conversations? Very excited to share the new MTRAG benchmark we have released! Data: github.com/ibm/mt-rag-ben… Paper: arxiv.org/abs/2501.03468 Danish Contractor Kshitij Fadnis Chulaka Gunasekara Vraj Shah Lucian Popa