jtsaw

I’d say it’s more like Waymo’s world model. The main actor uses a latent vector representation of the state of the game to make decisions. This latent vector at train time is meant to compress a bunch of useful information about the game. So while you can’t really understand the actual latent vector that represents state, you do know it encodes at least the state of the game.

This world model stuff is only possible in environments that are sandboxed. Ie you can represent the state of the world in an and have a way of producing the next state given a current state and action. Things like Atari games, robot simulations, etc

I'm not sure they are highly correlated. A committee uses the same LLM with the same input context to generate different outputs. Given the same context LLMs should produce the same next token output distribution (assuming fixed model parameters, temperature, etc). So, while tokens in a specific output are highly correlated, complete outputs should be independent since they are generated independently from the same distribution. You are right they are not iid but the calculation was just a simplification.

yea. We're definitely concerned about hallucinations and are using a variety of techniques to try and mitigate it (there's some existing discussion here, but using committees and sub-agents responsible for smaller tasks has helped).

What's helped the most, though, is using cluster information to back up decision making. That way we know the data it's considering isn't garbage, and the outputs are backed up by actual data.

one thing we're experimenting to help with the hallucinations/error rate issue is using a committee framework where we take a majority vote.

If the error rate of 1 expert is 5%, then for a committee of 10 experts, the probability a majority of the committee errors is around 0.00276% (binomial distribution with p=0.05). For 10 steps, this would be an error rate of 0.0276%

which website doesn't load for you?

The product will automatically execute runbooks for you. So far we've focused on using runbooks customers already have, since they know they work for them. We've also added the ability to turn of automatic execution for cases like a suggested runbook, so the customer can make any edits if necessary before approving it to be executed automatically.

Yea, this is a big challenge for us. We're using a variety of strategies to make sure hallucinations are rare, but that's why we're also committed to not executing actions that modify your cluster unless explicitly specified in a runbook

yea, we'd like to actually create these issues on a real cluster, but we couldn't figure out a good way of doing it at scale. The best alternative that we could think of was using an LLM that knows the root cause and could hopefully simulate outputs of commands consistently. Let us know if you have other ideas, we're always looking for ways to improve it.

We're not using our AI Agent to determine if your answers are correct or not. We're just using something off the shelf.

try to be as detailed as possible. With text sometimes it's hard to determine how close your answer was to the correct one.

hmmm, not sure where the latency is coming from, might need to up the resources.

We're hoping to expand the dataset into other SRE tasks that aren't kubernetes related. It's just what we've focused on for right now.

yeah that makes sense. While the tech develops, our focus is on building an AI agent that can determine the root cause of an issue, which itself is an import step in fixing things.