Searching...
12 results for “ai behavior”
if we could take at this point a bit of a tangent and talk about education and learning. If you're somebody listening to this who's a smart person interested in programming, interested in AI, so I presume building something from scratch is a good beg
2:12:45 / 2:13:46
from transcript2:12:45 – 2:13:46
There's for example, the DCLM datasets, where they also train the classifier, but not to detect educational content. Instead, they trained it on open HERMIS datasets, which is a datasets for instruction tuning. And also the ExplainLikeIM5 subrredit.
14:11 / 15:16
from transcript14:11 – 15:16
And the final stage is much newer and will link to what is done in r one and these reasoning models is, I think, OpenAI's name for this. They had this new API in the fall, which they called the reinforcement fine tuning API. This is the idea that you
27:37 / 28:47
from transcript27:37 – 28:47
But when it comes to mission critical applications, if you're doing something in accounting, if you're doing customer support, obviously if you're a developer and you're pushing code to a server, it needs to be a lot more reliable. And and that's why
3:46 / 4:50
from transcript3:46 – 4:50
to be more leveraged. Today, the dominant paradigm is reinforcement learning. And oftentimes, these agents are trained through reinforcement learning, where you'd build what's called an RL environment, which is like a mini world model for business. Y
8:07 / 9:58
from transcript8:07 – 9:58
You just And then you build up some intuition like this this mixture of expert works, and then this implementation of MLA works. Key hyperparameters like learning rate and regularization and things like this, and you find the regime that works for yo
56:55 / 58:01
from transcript56:55 – 58:01
With agents, and let's define an agent. I mean, different people define agents in different ways. I would define an agent as something that's capable of taking action in the real world or in the physical world. Something that's executing a multi step
7:27 / 8:33
from transcript7:27 – 8:33
Please. So someone would like to take the the LLM or as a new version of them sometimes called LRMs, large reasoning models, which is really a misnomer, but system like, you know, o one, o three, r one, and try to get them to behave through guardrail
4:32 / 5:34
from transcript4:32 – 5:34
So for example, for them, they always see the generations on either source code or raw text documents, and then they rewrite them to make sure they're easier to generate instructions from. And then they use that for their like instruction data genera
15:36 / 16:48
from transcript15:36 – 16:48
And here are just, some, some examples, to show, you know, what we we tried at scale. We basically trained a 1.7 billing model to try to do some, text completion, to try to beat LMs at their own game, which we kinda failed, but, we we gave it a hones
3:56:57 / 3:58:02
from transcript3:56:57 – 3:58:02
So at a high level, we train models that predict how well someone will perform on a job better than a human can. So similar to how a human would review a resume, conduct an interview, and decide who to hire, we automate all of those processes with LL
0:48 / 1:50
from transcript0:48 – 1:50
adjusts based on the new incoming information, does so continually and rapidly and frequently and so on. And then the thing you mentioned on the other side of it is generally, it will be referred to as in context learning. As you learn stuff, there's
2:55:27 / 2:56:34
from transcript2:55:27 – 2:56:34
Have a podcast?
Run your podcast through Clypt
Get ranked clips, hooks, and ready-to-post copy from your own episodes. Free to try.