AI Training Datasets

NOTICE: DO NOT USE THIS DATASET FOR ANY COMMERCIAL OR AI TRAINING PURPOSES. UNAUTHORIZED REDISTRIBUTION IS STRICTLY PROHIBITED. 本データセットは、金融庁のEDINETが公開しているXBRLデータを著作者が独自に解析・蓄積したものです。 - 本データセットは Gated Dataset です。 - データの取得を希望する場合は、Hugging Faceアカウントでログインし、アクセスリクエストを行ってください。 - 2026年5月2日：ライセンスをMITから独自規約に変更し、アクセス制限（Gated/Manual Approval）を導入しました。

Growing, publicly downloadable video-backed rollout dataset for the tracemarkov ablation arm at training step 500, seed 20260921. Each selected frozen observational prefix has exactly one cached continuation in each of two routes: null-intervention observational and diagnostic do(X=1). Prefix IDs are sorted once within each (W,Z) stratum and remain balanced as the cohort grows. The direct route of an observationally trained arm is a diagnostic; it does not replace the paper's observational importance-weighted estimator. Sampling note. Rollouts use a growing key-value cache: every generated frame attends to the whole history since the prefix, up to 280 frames, at absolute temporal positions.…

This repository is a private migration archive, not a public dataset release or a single uniform loaddataset table. Companion code, detailed recovery manual and sanitized research/conversation history: Start with GitHub RESTORE.md, CODEXHANDOFF.md, and docs/RESEARCHHISTORY.md. Download only the explicit files in migration/artifactsmanifest.json, using the frozen HF commit recorded in the companion GitHub migration/artifactrelease.json. Contents: eight consecutive open research stages, base examples, lifetime fitting, calibration, generated controlled edits, split manifests, predictions, training checkpoints, all failed attempts/configuration logs, API usage estimates, tables and reports…

Dataset · Other

seatgeek-dataset

Rebrowser.net

prettyname: "SeatGeek Events & Ticket Listings Dataset" - open-data - data-science - web-scraping - scraper - crawler - data-collection - tickets - seatgeek - events - concerts - sports - ticket-prices - deal-score sizecategories: taskcategories: - other - configname: events datafiles: - configname: event-listings datafiles: - configname: performers datafiles: - configname: venues datafiles: Daily sample of SeatGeek events, ticket listings, performers, and venues with Deal Score ratings, section-level seating, delivery types, and cross-platform IDs. This dataset is a preview sample of the SeatGeek dataset published by Rebrowser. If you're doing academic research, you may be eligible for…

Сборка: 2026-10-14 03:40 UTC Файл links.db — SQLite: таблицы entities, mentions, edges. Запросы: python scripts/entitylinks.py search --db links.db --id. Источники: ЄДР, ПДВ, ЄДРСР, а також person-реестры (wanted, missing, notaries, courtexperts, arbitrationmanagers, advocates, auditors, debtors). Точная входная provenance: graphprovenance.json содержит URL, локальный путь, размер и SHA-256 каждого успешно скачанного файла. Рёбра identity/nameresolve — консервативное разрешение по ФИО. обезличивание источника сохраняется (nodeanonymization).

Pages and images of 61 catalog websites (40 topic categories), collected live from the sites' public origins. Candidate URLs are the sites' complete Common Crawl URL inventories (all 128 collections, legacy indexall selection) plus links discovered on saved pages. raw/.origin.html.gz — the HTML exactly as served (gzip). pages/.html.gz — a passive, cleaned HTML rendition (no scripts, styles or forms; ordered media), with the page's images localized once they were downloaded. documents/.json.gz — the manifest: URL, capture time, HTTP metadata, provenance (Common Crawl record or discovering page), quality label, title, ordered media list with each image's local path and SHA-256. Images are…

Dataset

loki-expLLM

squad (FR)

This dataset contains interaction logs from the Loki project, a research initiative focused on creating scalable, LLM-powered honeypots. The data captures requests made to three different honeypot experiments (Exp0, Exp1, and Exp2), each with a unique architecture and purpose. The primary goal of this dataset is to provide researchers and security professionals with real-world examples of interactions with different honeypot strategies. The interactions include legitimate-looking traffic, automated scanning, and various exploitation attempts such as SQL Injection (SQLi), Cross-Site Scripting (XSS), Remote Code Execution (RCE), and fuzzing. This dataset can be used for a variety of tasks in…

Dataset

komorebi-painter-teachers

Chinmay Harshavardhan Karkar

This dataset stores the 40 fixed COCO128 reference JPEGs used by the painter teacher benchmark. The source metadata and per-image provenance are in refs.json; SHA-256 hashes are in reference-manifest.json. The source records do not establish per-image licensing, so this dataset card does not assert a blanket image license. Generated benchmark results will be uploaded under runs/ with their own manifests after each run.