Base Labs, Hugging Face, and Goodfire AI announced a partnership to develop transparency standards and monitoring infrastructure for open-weight models, addressing risks associated with 'abliteration' techniques. The initiative aims to embed safety protocols directly into the training and deployment lifecycle of open-source models rather than applying them as post-hoc fixes. This development signals an industry effort to create verifiable safety baselines for open models that are currently accessible on major hosting platforms.
89 other tracked developments touching the same claims.