Project: RL1/RL2 (obsolete)
Collection
Older models that are no longer useful for anything in RL1 or RL2, or are now unused as experimentation discontinued. • 16 items • Updated
YAML Metadata Warning:empty or missing yaml metadata in repo card
Check out the documentation for more information.
OBSOLETE
These models were part of the experiment to cut the antenna off the butterfly. They are obsolete. Do not use. Have been replaced with these models.
Wandb: https://wandb.ai/devinterp/jaxgmg_banlist
Models trained with restrictions: Environment is (1-alpha) chance of cheese_always_in_the_corner, and alpha chance of the environment:
- banlist_c21_c12: Cheese/mouse can be anywhere, except cheese = (2,1) or cheese = (1,2)
- Only 4/20 runs learn optimla behaviour everywhere
- whitelist_c21_mrow1_c12_mcol1: Cheese/mouse can only be in
- cheese = (2,1), mouse uniformly placed in row 1
- cheese = (1,2), mouse uniformly placed in col 1
- Very noisy training, no clear phases, 0/20 policies learn optimal View action probs here
Models
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support