ryoungj/bootstrap-latent-thought-data
This dataset is associated with the paper Reasoning to Learn from Latent Thoughts. It contains data used for pretraining language models with a focus on improving data efficiency by modeling and inferring latent thoughts underlying the text generation process, such as on reasoning-intensive math corpus. An expectation-maximization algorithm is developed for models to self-improve their self-generated thoughts and data efficiency.
| Author / organization | ryoungj |
|---|---|
| License (as declared) | apache-2.0 |
| Downloads | 1.1K (as reported by Hugging Face) |
| Likes | 4 |
| Created | 20 March 2025 |
| Last updated at source | 31 March 2025 |
Popularity, last 3 days
1,061 now (+99 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.