alunxu/spatial-memory-checkpoints

README: strip idea-revealing framing; load-only

README: clarify frames-per-ckpt mapping (blind=10.06M, sighted=5.0M)

Add full training trajectory for foveated_logpolar (50 ckpts, post-retrain)

Add full training trajectory for uniform (50 ckpts, post-retrain)

Add full training trajectory for foveated (50 ckpts, post-retrain)

Add full training trajectory for coarse (50 ckpts, post-retrain)

Add full training trajectory for blind (35 ckpts, post-retrain)

Clean stale ckpts + minimal README (post-retrain release in flight)

Update to post-retrain hp-consistent 5-condition checkpoints; minimal README

Minimal README: load snippet only

Minimize README (protect idea)

Initial upload: 4-cond DD-PPO PointNav checkpoints + README

initial commit