Dataset · Hugging Face

CollinsStreet/ReLaDrift-data

Name: ReLaDrift-data
Creator: CollinsStreet
License: mit
Keywords: task_categories:text-generation, language:en, license:mit, size_categories:1M<n<10M, format:json, modality:tabular, modality:text, library:datasets, library:pandas, library:polars, library:mlcroissant, arxiv:2610.23225, region:us, math, code

ReLaDrift GSM8K training data GSM8K training data of ReLaDrift: 1,588,781 TinyGSM questions, each with 12 distinct executable programs generated by Qwen3.5-9B (selected from 32 samples; answers are not checked). Code: Yuki-Ots/ReLaDrift. File Contents selected.jsonl question, the 12 codes and their answers, row_id, source_index, sample_indices, lexical_signatures canonical_row.npy Row of each question in the TinyGSM token cache (uint64) train512.json The 512… See the full description on the dataset page: https://huggingface.co/datasets/CollinsStreet/ReLaDrift-data.

View on Hugging FaceReport an issue

Author / organization	CollinsStreet
License (as declared)	mit
Downloads	0 (as reported by Hugging Face)
Likes	0
Created	21 October 2026
Last updated at source	21 October 2026

mathcode

Embed this entry

<iframe src="https://silytics.si/embed/item/3f2f23d0-cff1-43fc-b98c-bfda00f6d852" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

Connected in the knowledge graph

CollinsStreet/ReLaDriftModel · used to train
ReLaDrift: Recurrent Language Generation via Drifting for Test-Time ScalingPaper · cites

Explore this entry in the graph

Related papers

ReLaDrift: Recurrent Language Generation via Drifting for Test-Time ScalingPaper

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 21 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.