ive been trying to figure out what I can and cannot turn into a classification problem.

Homeby Sid2026-09-17

ive been trying to figure out what I can and cannot turn into a classification problem. comparing a couple different wordle stratergies is actually a good basic case study. I think Jev lowkey is pretty greedy (when putting it in 'multi-turn' type setups - like letter by letter guessing words). Its kind of obivous but pretty interesting https://x.com/sid19arya0/status/2100679099723223341?s=20

Links

Post by @sid19arya0x.comJev @typesafeai playing wordle - with different stratergies is a pretty good case study in what Jev can and cannot do. Guess letter-by-letter -> very greedy All 5 letters at once -> spams SEEEE (repeatedly) Guess the who

open source discussion

More in Home

I tried to create a latent context layer that is a warmed up version of a small LDA Topic model that is injectable into the context window via memories or system prompts.I made some CVSS classification scripts that use Jev to generate a CVSS vector for a given software vulnerability description.Jev can be used for profiling players, which is great for organizing playtests and checking balance changes across different player groups.