TypeSafe's new decision model, Jev, has been shown playing games, so I tried it on Manic Miner. It works - but there is a big question - is Jev actually doing anything?

2 points•iamflimflam1•about 11 hours ago•3 comments•

3 comments

verdvermabout 10 hours ago
Did you watch Jev play pokemon? It showed interesting patterns with that game too. I'm going the llm route and having it play my first RPG. Turn based so latency is acceptable, prefer the smarts for this one.

I've been pondering if I even bother with map views, or just provide navigable locations as aist like Jev-Pomemon. It'll end up having both options I'm sure, since I'm using the project to explore harness evals.

iamflimflam1about 10 hours ago
Interesting idea - so instead of "which move do you want to make?" it's "which location do you want to go to?".
verdvermabout 10 hours ago
yeah, that was how jev-pokemon worked

since mine is more of a harness/eval framework, I'm going to experiment with different toolsets and context constructions

It's definitely easier to skip that part and just use A* for navigation, but I also want to see it explore and learn the map, through more basic moves. Right now I'm not exposing direct button presses to the agent.

Read the full thread on Hacker News →

Related stories