Synthetic Pre-pretraining Survives Scale, but Not as a Grammatical Prior
Paper • 2609.39827 • Published • 12
Dataset Viewer
The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
Synthetic Pre-pretraining Datasets
This dataset contains the pre-processed datasets used in the paper Synthetic Pre-pretraining Survives Scale, but Not as a Grammatical Prior.
- Paper: Hugging Face Paper Page
- Code: GitHub Repository
- Project Page: Hugging Face Project Page
The data includes pre-pretraining (PPT) tasks such as k-Shuffle Dyck, Set, NCA, and MP-Struct Core, as well as control baselines and pre-training corpora (C4, SmolLM3, Olmo3, Marin). All datasets are available under the verify-ppt organization.
For full preprocessing, training, and evaluation instructions, refer to the GitHub repository.
- Downloads last month
- 11