update readme

README.md

@@ -1,8 +1,10 @@
 # jaxgmg2_shared_init
 
-20 RL agent checkpoints studying the effect of shared initialization. Two base models (run_ids 19 and 27
-from jaxgmg2_3phase_optim_state) are each used as a shared starting point, then independently continued
-from checkpoint 0 (fresh optimizer state) with alpha=1.0 across 10 different random seeds each.
+20 RL agent checkpoints studying the effect of shared initialization. Two base models
+(`al_1.0_g_0.98_id_19_seed_981019` and `al_1.0_g_0.98_id_27_seed_981027` from
+[jaxgmg2_3phase_optim_state](https://huggingface.co/timaeus/jaxgmg2_3phase_optim_state))
+are each used as a shared starting point, then independently continued from checkpoint 0
+(fresh optimizer state) with alpha=1.0 across 10 different random seeds each.
 
 **WandB:** https://wandb.ai/devinterp/jaxgmg2_shared_init
 
@@ -48,7 +50,7 @@ Checkpoints are named `al_1.0_g_0.98_id_{run_id}_shared_init_seed_{seed}`.
 See `train.yaml` in this repository. Run with:
 
 ```bash
-make run projects/rl/experiments/shared_init/jobs/train.yaml
+timaeus run train.yaml
 ```
 
 from the [timaeus monorepo](https://github.com/timaeus-research/timaeus).