Dylan Patel

And that model, which the weights are available, has no human preferences added into the post-training. The DeepSeq R1 full model has some of this human preference tuning, this RLHF, after the reasoning stage. But the very remarkable thing is that you can get these reasoning behaviors And it's very unlikely that there's humans writing out reasoning chains.

And that model, which the weights are available, has no human preferences added into the post-training. The DeepSeq R1 full model has some of this human preference tuning, this RLHF, after the reasoning stage. But the very remarkable thing is that you can get these reasoning behaviors And it's very unlikely that there's humans writing out reasoning chains.

And that model, which the weights are available, has no human preferences added into the post-training. The DeepSeq R1 full model has some of this human preference tuning, this RLHF, after the reasoning stage. But the very remarkable thing is that you can get these reasoning behaviors And it's very unlikely that there's humans writing out reasoning chains.

It's very unlikely that they somehow hacked OpenAI and they got access to OpenAI 01's reasoning chains. It's something about the pre-trained language models and this RL training where you reward the model for getting the question right. And therefore, it's trying multiple solutions and it emerges this chain of thought.

It's very unlikely that they somehow hacked OpenAI and they got access to OpenAI 01's reasoning chains. It's something about the pre-trained language models and this RL training where you reward the model for getting the question right. And therefore, it's trying multiple solutions and it emerges this chain of thought.

It's very unlikely that they somehow hacked OpenAI and they got access to OpenAI 01's reasoning chains. It's something about the pre-trained language models and this RL training where you reward the model for getting the question right. And therefore, it's trying multiple solutions and it emerges this chain of thought.

I think it's good to recap AlphaGo and AlphaZero because it plays nicely with these analogies between imitation learning and learning from scratch. So AlphaGo, the beginning of the process was learning from humans where they had, they started the first, this is the first expert level Go player or chess player in DeepMind series of models where they had some human data.

I think it's good to recap AlphaGo and AlphaZero because it plays nicely with these analogies between imitation learning and learning from scratch. So AlphaGo, the beginning of the process was learning from humans where they had, they started the first, this is the first expert level Go player or chess player in DeepMind series of models where they had some human data.

I think it's good to recap AlphaGo and AlphaZero because it plays nicely with these analogies between imitation learning and learning from scratch. So AlphaGo, the beginning of the process was learning from humans where they had, they started the first, this is the first expert level Go player or chess player in DeepMind series of models where they had some human data.

And then why it is called AlphaZero is that there was zero human data in the loop. And that change to AlphaZero made a model that was dramatically more powerful for DeepMind. So this remove of the human prior, the human inductive bias, makes the final system far more powerful. We mentioned Bitter Lesson hours ago, and this is all aligned with this.

And then why it is called AlphaZero is that there was zero human data in the loop. And that change to AlphaZero made a model that was dramatically more powerful for DeepMind. So this remove of the human prior, the human inductive bias, makes the final system far more powerful. We mentioned Bitter Lesson hours ago, and this is all aligned with this.

And then why it is called AlphaZero is that there was zero human data in the loop. And that change to AlphaZero made a model that was dramatically more powerful for DeepMind. So this remove of the human prior, the human inductive bias, makes the final system far more powerful. We mentioned Bitter Lesson hours ago, and this is all aligned with this.

And then there's been a lot of discussion in language models. This is not new. This goes back to the whole QSTAR rumors, which if you piece together the pieces is probably the start of OpenAI figuring out its O1 stuff when last year in November, the QSTAR rumors came out.

And then there's been a lot of discussion in language models. This is not new. This goes back to the whole QSTAR rumors, which if you piece together the pieces is probably the start of OpenAI figuring out its O1 stuff when last year in November, the QSTAR rumors came out.

And then there's been a lot of discussion in language models. This is not new. This goes back to the whole QSTAR rumors, which if you piece together the pieces is probably the start of OpenAI figuring out its O1 stuff when last year in November, the QSTAR rumors came out.

There's a lot of intellectual drive to know when is something like this going to happen with language models, because we know these models are so powerful and we know it has been so successful in the past. And it is a reasonable analogy that this new type of reinforcement learning training for reasoning models is when the door is open to this.

There's a lot of intellectual drive to know when is something like this going to happen with language models, because we know these models are so powerful and we know it has been so successful in the past. And it is a reasonable analogy that this new type of reinforcement learning training for reasoning models is when the door is open to this.

There's a lot of intellectual drive to know when is something like this going to happen with language models, because we know these models are so powerful and we know it has been so successful in the past. And it is a reasonable analogy that this new type of reinforcement learning training for reasoning models is when the door is open to this.

We don't yet have the equivalent of turn 37, which is the famous turn where the DeepMind's AI playing ghost dumped Lee Sedol completely. We don't have something that's that level of focal point, but that doesn't mean that the approach to technology is different and the impact of the general training. It's still incredibly new. What do you think that point would be?

We don't yet have the equivalent of turn 37, which is the famous turn where the DeepMind's AI playing ghost dumped Lee Sedol completely. We don't have something that's that level of focal point, but that doesn't mean that the approach to technology is different and the impact of the general training. It's still incredibly new. What do you think that point would be?