transformer-shortest-paths

Experimentally evaluating transformers' generalization on a synthetic task (mirror)

Commits at 55f2c364c6d9b9d2d9549a55deac7e1416eabc02

  1. 55f2c364 Add first draft of blog post SIPB authored at SIPB comitted at
  2. 3982e8fa Make the target embedding learnable SIPB authored at SIPB comitted at
  3. c8c1abdd Better plots SIPB authored at SIPB comitted at
  4. 072c87a3 New embeddings and readout SIPB authored at SIPB comitted at
  5. 68887a78 Decrease num layers SIPB authored at SIPB comitted at
  6. eab86fac MORE LAYERS SIPB authored at SIPB comitted at
  7. e953b375 Use bfloat16 SIPB authored at SIPB comitted at
  8. e460b0ae idk what this does but it's the latest version of TSP SIPB authored at SIPB comitted at
  9. 5654942c Generate data on the fly SIPB authored at SIPB comitted at
  10. 1fe0e023 Minor README tweaks SIPB authored at SIPB comitted at
  11. d3e7c909 updates Alek Westover authored at Alek Westover comitted at
  12. 461d3543 Delete unused images dir SIPB authored at SIPB comitted at
  13. 11d377f6 Condense proposal into 1 page sipb authored at sipb comitted at
  14. 26b5a9d8 wrote project proposal @anthowan Anthony can you make sure it looks good? Alek Westover authored at Alek Westover comitted at
  15. 49059938 Tasks 1 through 4 SIPB authored at SIPB comitted at
  16. 8c2dc28c update Alek Westover authored at Alek Westover comitted at
  17. 56230674 fixed some bugs Alek Westover authored at Alek Westover comitted at
  18. 15e149cb update Alek Westover authored at Alek Westover comitted at
  19. 5f733f89 add readme Alek Westover authored at Alek Westover comitted at
  20. 0b633a51 upload Alek Westover authored at Alek Westover comitted at
  21. caf0dfc4 add Alek Westover authored at Alek Westover comitted at