1/ Our models are becoming more capable in biology and we expect upcoming models to reach ‘High’ capability levels as defined by our Preparedness Framework. 🧵
Johannes Heidecke
137 posts
- 1 / Today we launched gpt-5, finishing a huge week for OpenAI. We’ve raised the bar for safety in both open and closed models. With gpt-oss and gpt-5, we introduced meaningful capability advancements with rigorous, industry-leading safeguards and safety testing.
- Safety is a core focus of our open-weight model’s development, from pre-training to release. While open models bring unique challenges, we’re guided by our Preparedness Framework and will not release models we believe pose catastrophic risks.TL;DR: we are excited to release a powerful new open-weight language model with reasoning in the coming months, and we want to talk to devs about how to make it maximally useful: openai.com/open-model-fee… we are excited to make this a very, very good model! __ we are planning to
- 🧵Today we’re sharing more details about improvements of the default GPT-5 model in responding to sensitive conversations around potential mental health emergencies and emotional reliance. These changes reflect the careful work of many teams within OpenAI and close consultationEarlier this month, we updated GPT-5 with the help of 170+ mental health experts to improve how ChatGPT responds in sensitive moments—reducing the cases where it falls short by 65-80%. openai.com/index/strength…
- 1/ Safety is core to every model we build at OpenAI. As we deploy GPT-4.1 into ChatGPT, we want to share some insights from our safety work. 🧵
- Proud to share our work on Deliberative Alignment openai.com/index/delibera… with a special shoutout to @Melodyguan who led this work. Deliberative Alignment trains models to reason over relevant safety and alignment policies to forge their responses.
- OpenAI is nothing without its people
- i love the openai team so much
- Replying to @JoHeidecke3/ Today, we are sharing more details on what we’re doing to mitigate this risk in our deployments, and some ideas for researchers, governments, and the world at large to accelerate our overall readiness.
- Replying to @JoHeidecke2/ This will enable and accelerate beneficial progress in biological research, but also - if unmitigated - comes with risks of providing meaningful assistance to novice actors with basic relevant training, enabling them to create biological threats.
- Using DALL·E 2 to imagine what weirdness Hieronymus Bosch might have painted if he lived today 🎨 #dalle #dalle2 #outpainting
Inpainting with DALL·E 2 is super fun. With some ingenuity, you can create arbitrarily large artwork like the murals shown below – which I assume are the largest #dalle-produced images created so far. - Open models can unlock huge benefits, and like any powerful technology, they carry misuse risks. Once the weights are released, there’s no pulling them back. This is why safety testing matters even more here. 1/Today we release gpt-oss-120b and gpt-oss-20b—two open-weight LLMs that deliver strong performance and agentic tool use. Before release, we ran a first of its kind safety analysis where we fine-tuned the models to intentionally maximize their bio and cyber capabilities 🧵
- Very proud of all the safety work we've done for o1 & new research directions of making our models safer and more aligned 🍓🥽
- Exciting work from OpenAI interpretability team! :)Excited to share our latest work on untangling language models by training them with extremely sparse weights! We can isolate tiny circuits inside the model responsible for various simple behaviors and understand them unprecedentedly well. openai.com/index/understa…
