Smith, Will
To Err is AI : A Case Study Informing LLM Flaw Reporting Practices
McGregor, Sean, Ettinger, Allyson, Judd, Nick, Albee, Paul, Jiang, Liwei, Rao, Kavel, Smith, Will, Longpre, Shayne, Ghosh, Avijit, Fiorelli, Christopher, Hoang, Michelle, Cattell, Sven, Dziri, Nouha
In August of 2024, 495 hackers generated evaluations in an open-ended bug bounty targeting the Open Language Model (OLMo) from The Allen Institute for AI. A vendor panel staffed by representatives of OLMo's safety program adjudicated changes to OLMo's documentation and awarded cash bounties to participants who successfully demonstrated a need for public disclosure clarifying the intent, capacities, and hazards of model deployment. This paper presents a collection of lessons learned, illustrative of flaw reporting best practices intended to reduce the likelihood of incidents and produce safer large language models (LLMs). These include best practices for safety reporting processes, their artifacts, and safety program staffing.
OLMo: Accelerating the Science of Language Models
Groeneveld, Dirk, Beltagy, Iz, Walsh, Pete, Bhagia, Akshita, Kinney, Rodney, Tafjord, Oyvind, Jha, Ananya Harsh, Ivison, Hamish, Magnusson, Ian, Wang, Yizhong, Arora, Shane, Atkinson, David, Authur, Russell, Chandu, Khyathi Raghavi, Cohan, Arman, Dumas, Jennifer, Elazar, Yanai, Gu, Yuling, Hessel, Jack, Khot, Tushar, Merrill, William, Morrison, Jacob, Muennighoff, Niklas, Naik, Aakanksha, Nam, Crystal, Peters, Matthew E., Pyatkin, Valentina, Ravichander, Abhilasha, Schwenk, Dustin, Shah, Saurabh, Smith, Will, Strubell, Emma, Subramani, Nishant, Wortsman, Mitchell, Dasigi, Pradeep, Lambert, Nathan, Richardson, Kyle, Zettlemoyer, Luke, Dodge, Jesse, Lo, Kyle, Soldaini, Luca, Smith, Noah A., Hajishirzi, Hannaneh
Language models (LMs) have become ubiquitous in both NLP research and in commercial product offerings. As their commercial importance has surged, the most powerful models have become closed off, gated behind proprietary interfaces, with important details of their training data, architectures, and development undisclosed. Given the importance of these details in scientifically studying these models, including their biases and potential risks, we believe it is essential for the research community to have access to powerful, truly open LMs. To this end, this technical report details the first release of OLMo, a state-of-the-art, truly Open Language Model and its framework to build and study the science of language modeling. Unlike most prior efforts that have only released model weights and inference code, we release OLMo and the whole framework, including training data and training and evaluation code. We hope this release will empower and strengthen the open research community and inspire a new wave of innovation.