Openai now reveals more than the thought process of O3-Mini Model

In response to pressure from rivals including Chinese AI company DeepseekOpenai changes the way his most new AI model, O3-MiniHe communicates his step -by -step process of Thought.

Thursday, Openai declared This free and paid ChatgPT users, a Chatbot platform powered by AI, will see an updated “thought chain” that shows more than the model’s “reasoning” steps and how it came to answers to questions. PREMIUM CHATGPT subscribers that use O3-Mini in the High Refaging configuration will also see this updated reading, according to Openai.

“We present an updated (chain of thought) for the O3-Mini designed to make it easier for people to understand how the model thinks,” a spokesman for Openai told TechCruch via email. “With this update you will be able to follow the model’s reasoning by giving you more clarity and confidence in its answers.”

Models for reasoning like O3-Mini are thoroughly checked before giving results that helps them Avoid some of the trails which usually discard models. The compromise is that the models of reasoning take a little longer to get to solutions-usually seconds to minutes longer.

R1 model of DeepseekThe “Reflections” model on the O3-Mini lines reveals its full thought process, which many AI researchers claim to be the preferred approach. In addition to facilitating the model easier to study, the steps of reasoning provide a better user experience in certain situations, helping to indicate when the model can be on the right – or wrong – a song.

Openai had decided not to show the full steps to think about O3-Mini and his predecessors, O1 and O1-Mini, partly for competitive reasons. Instead, the users saw only summaries of the steps of reasoning – summons that were sometimes wrongS

When we informed the people of 🍓 before the release of the O1-Preview, the view of the live basket was usually a “aha” moment for them, which made it clear that it would be a great job. These are not the harsh beds, but this is a big step closer and I’m glad we can share this experience with the world. https://t.co/72zpprhmfk

– Noam Brown (@polynoamial) February 6, 2025

Openai still does not show the full steps to think of O3-Mini, but the company said it “found balance”: O3-Mini could “think freely” and then organize its “thoughts” in more detailed generalizations.

“To improve clarity and safety, we added an additional step after processing, in which the model views the raw chain of thought, eliminating any dangerous content and then simplifies any complex ideas,” continued Openai spokesman. “In addition, this step after processing allows users who are not English to receive the mental chain in their native language, creating a more accessible and friendly experience.”

In a Reddit ama last weekKevin Weil, Chief Product Officer of Openai, hinted that the change is coming.

“We are working on showing a bunch more than we show today – (showing the model thought process) it will be very, very soon,” he said. “TBD for everyone – displaying the whole chain of thought leads to competitive distillation, but we also know that people (at least power users) want it, so we will find the right way to balance it.”

TechCrunch has a newsletter focused on AI! Sign up here To get it in your incoming mail every WednesdayS