Estimating worst case frontier risks of open weight LLMs
Research paper introducing malicious fine-tuning (MFT) methodology to estimate worst-case frontier risks when releasing open-weight LLMs in biology and cybersecurity domains.
Excerpt
In this paper, we study the worst-case frontier risks of releasing gpt-oss. We introduce malicious fine-tuning (MFT), where we attempt to elicit maximum capabilities by fine-tuning gpt-oss to be as capable as possible in two domains: biology and cybersecurity.
Read at source: https://openai.com/index/estimating-worst-case-frontier-risks-of-open-weight-llms