Card: CC BY 4.0
Browse files
README.md
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
---
|
| 2 |
-
license:
|
| 3 |
library_name: pytorch
|
| 4 |
tags:
|
| 5 |
- recursive-reasoning
|
|
@@ -44,5 +44,7 @@ reproducibility only; the same splits can be rebuilt from those releases with
|
|
| 44 |
`python scripts/data/build_eval_cohorts.py` in the code repository. The DF split of each
|
| 45 |
benchmark is the set of Test puzzles the frozen backbone fails to solve deterministically.
|
| 46 |
|
| 47 |
-
The checkpoints are released under
|
| 48 |
-
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
license: cc-by-4.0
|
| 3 |
library_name: pytorch
|
| 4 |
tags:
|
| 5 |
- recursive-reasoning
|
|
|
|
| 44 |
`python scripts/data/build_eval_cohorts.py` in the code repository. The DF split of each
|
| 45 |
benchmark is the set of Test puzzles the frozen backbone fails to solve deterministically.
|
| 46 |
|
| 47 |
+
The checkpoints and the evaluation splits are released under CC BY 4.0: credit digiLab
|
| 48 |
+
Solutions Ltd and cite the paper when you use them. The splits are derived from the HRM
|
| 49 |
+
releases, whose own terms apply to the underlying puzzles; `NOTICES.md` in the code
|
| 50 |
+
repository lists the licences of the datasets and of the backbone code.
|