datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
assetssnapshotsunienv_genesis_collection_assetsGenesisII
QVAC Genesis II: Expanding the Largest Multi-domain Educational Synthetic Dataset for Pre-training
📖 Read the full blog post on Hugging Face | 🔗 Genesis I Dataset
QVAC Genesis II is a major expansion of the largest publicly available education-focused synthetic dataset for LLM pre-training and reasoning-centric post-training. Building upon Genesis I, it adds 10 new educational domains and introduces a novel Option-Level Reasoning Analysis data generation method, totaling 86… See the full description on the dataset page: https://huggingface.co/datasets/qvac/GenesisII.adam-worldGenesisI
QVAC Genesis I: the Largest and Highest-Quality Multi-domain Educational Synthetic Dataset for Pre-training
QVAC Genesis I is a large-scale education-focused synthetic dataset (40.9B tokens; 31.8M rows) purpose-built for LLM pre-training and reasoning-centric post-training. It covers Mathematics, Physics, Biology, Medicine, and Logical Deduction across high-school and college/professional levels. Data are generated via a scalable learn-from-failures pipeline: seed → MCQs → model… See the full description on the dataset page: https://huggingface.co/datasets/qvac/GenesisI.tactile-genesis-heavy-mg-v2
Tactile Genesis heavy-cube pick-and-place (heavy_mg_v2)
272 success-filtered scripted-expert episodes (158 frames each, 60 Hz sim / 1 control step
per frame) of a Franka Panda picking up a randomized-mass grey cube (0.24–1.19 kg,
uniform-by-density; visually identical across masses) among 3 distractor cubes and placing
it in a basket, in the Genesis simulator.
The scripted expert is tactile-reactive: grip force ramps on sensed slip. Collected with
MimicGen-style stratified… See the full description on the dataset page: https://huggingface.co/datasets/jalal06/tactile-genesis-heavy-mg-v2.stallion-backup-genesis-rerun-quebec-20260713
Genesis × Rerun: the Bottle-Flip Data Flywheel
Idea → thousands of parallel physics trials → Rerun recordings & SQL catalog →
LeRobot dataset → trained policy → closed-loop sim eval. One prompt-sized idea
("a robot arm flips a water bottle"), driven all the way to a visuomotor policy,
with Rerun as the data backbone at every step.
Everything here is contact physics: a Franka Panda pinches the bottle's neck
under the cap lip (form closure), swings, opens its fingers mid-arc, and… See the full description on the dataset page: https://huggingface.co/datasets/Reza2kn/stallion-backup-genesis-rerun-quebec-20260713.fake-real-newsGenesis-knowledge
Genesis Knowledge
A structured symbolic knowledge dataset for the Genesis project.
This repository contains the editable knowledge source used by Genesis. It is designed to be compiled into a binary knowledge graph for symbolic reasoning systems.
The dataset contains concepts, properties, relationships, categories, and metadata stored in a simple, human-readable .kc format.
Purpose
Genesis Knowledge is intended to be a structured, expandable knowledge base that… See the full description on the dataset page: https://huggingface.co/datasets/XPDevs/Genesis-knowledge.genesis-stack-cube-2genesis-1k-ignorealexandria-aeternum-genesis
Alexandria Aeternum
Cognitive Nutrition for Foundation Models
10,097 provenance-verified artworks · C2PA Content Credentials · Full images + metadata · 29 structured columns · 4,000+ tokens each
Not scraped. Not auto-captioned. Translated from human knowledge. Cryptographically sealed.
MCP Access (2M+ Artworks) · Research Paper · Explore Full Archive · Scale With Us · Research Partnerships
MCP Access — AI Agent Marketplace
This complete 10… See the full description on the dataset page: https://huggingface.co/datasets/Metavolve-Labs/alexandria-aeternum-genesis.OS-Genesis-mobile-data🫶 If you are interested in our work or find this data helpful, please consider using the following .bib when referencing our paper:
@article{sun2024genesis,
title={OS-Genesis: Automating GUI Agent Trajectory Construction via Reverse Task Synthesis},
author={Sun, Qiushi and Cheng, Kanzhi and Ding, Zichen and Jin, Chuanyang and Wang, Yian and Xu, Fangzhi and Wu, Zhenyu and Jia, Chengyou and Chen, Liheng and Liu, Zhoumianze and others},
journal={arXiv preprint arXiv:2412.19723}… See the full description on the dataset page: https://huggingface.co/datasets/OS-Copilot/OS-Genesis-mobile-data.genesis-stack-cube-2GenesisGeo 📃Paper • 📚 Github • 🤖 Model
Dataset Summary
This is a large-scale synthetic dataset containing 21,799,134 geometric theorem proving problems with complete proof traces, with 2,964,316 of these problems involve auxiliary constructions in their proofs. Generated by the GenesisGeo framework built upon Newclid, this dataset extends the original AlphaGeometry methodology and provides a rich resource for training neuro-symbolic reasoning models in Euclidean geometry.… See the full description on the dataset page: https://huggingface.co/datasets/ZJUVAI/GenesisGeo.genesis-pick-cube-stateThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "franka",
"total_episodes": 209,
"total_frames": 41800,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 60,
"splits": {
"train": "0:209"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/alt250/genesis-pick-cube-state.doe-genesis-sealed-n2500
Demonstration receipts (n=2500)
These are demonstration envelopes from a filing. n=2500 is a demonstration number. Each row is one trajectory summary, not a 1 kHz pulse and not a timeseries. 15 banks (37,500 rows). Per-row proof_hash. File cryptographic_seal.
What a stranger sees if they cite this zip
They land on a grant-shaped shelf: 15 configs next to each other, mass/μ/booleans/proof_hash. No pulse. No 4×4 taxels. No sentence that this is the direction for… See the full description on the dataset page: https://huggingface.co/datasets/spiderpilot89/doe-genesis-sealed-n2500.genesis_self_collect_pick_red_cube_ee_delta_v21This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "panda",
"total_episodes": 92,
"total_frames": 49225,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 10,
"splits": {
"train": "0:92"},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/yananchen/genesis_self_collect_pick_red_cube_ee_delta_v21.polymarket-genesis-liquidity
Polymarket Genesis-Phase Liquidity — Dome L2 Order-Book Snapshots
Level-2 top-of-book snapshots for Polymarket outcome tokens, fetched via the
Dome API (/polymarket/orderbooks). Companion dataset to the paper
Opening Wide, Moving Fast: Spread Dynamics in Polymarket Prediction Markets
(MARBLE 2026).
Rows: 339,183,655
Collection window: Oct 2025 – Mar 2026
One file: data/dome_polymarket_2025-10_2026-03.parquet (~14 GB, zstd)
Grain: one row per (timestamp_ms, token_id… See the full description on the dataset page: https://huggingface.co/datasets/vpcapitano/polymarket-genesis-liquidity.genesis_joint_position_2000_20fpsThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "panda",
"total_episodes": 2007,
"total_frames": 423477,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 3,
"chunks_size": 1000,
"fps": 20,
"splits": {
"train": "0:2007"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/kaveh-kamali/genesis_joint_position_2000_20fps.genesis-hr-bench-rebuttal-full19-hrv2-eval-20260812
Genesis HR Bench — preliminary HR v2 train19 rebuttal evaluation
This is an immutable, sanitized snapshot of the currently available artifacts from:
hf19_eval5_act080000_full95_fast_20260812T0236Z
It contains per-episode videos, state traces, result JSON, validation records, task/model manifests, policy-server logs, and Slurm logs for five policies across the 19 HR v2 tasks used for this rebuttal evaluation. It also includes both focused seed-7 skillet diagnostic pilots.… See the full description on the dataset page: https://huggingface.co/datasets/zimplex/genesis-hr-bench-rebuttal-full19-hrv2-eval-20260812.genesis_pick_cube_vr_absThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "panda",
"total_episodes": 50,
"total_frames": 11920,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 10,
"splits": {
"train": "0:50"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/yananchen/genesis_pick_cube_vr_abs.yao-genesis-audiopick-cube-state-genesisThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "franka",
"total_episodes": 209,
"total_frames": 41800,
"total_tasks": 1,
"total_videos": 0,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 60,
"splits": {
"train": "0:209"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/jadechoghari/pick-cube-state-genesis.functional-grasp-demos
Selected Grasp Demos
Contents
Each case folder contains:
grasp_data.npz — Top-1 ranked grasp (pre_grasp_dofs, grasp_target_dofs, reward, z_lift)
image_grasp.png — AI-generated grasp image (input to perception pipeline)
debug_retarget.png — WujiHand retarget visualization (if available)
Cases
Case
Object
Scale
Reward
z_lift
paper_coffee_cup_rot090__grasp_06
Paper Coffee Cup
1.0
0.742
0.196
paper_cup_8_oz_rot000__grasp_06
Paper Cup 8 Oz
1.0
0.696… See the full description on the dataset page: https://huggingface.co/datasets/Genesis-Intelligence/functional-grasp-demos.Genesis-PickLift-Easy-autogen-1024
Genesis HIL-SERL Demonstrations
LeRobotDataset v3.0 demonstrations for reproducing imitation-learning work and pretraining VLA
models.
Repository: JinbiaoZhu/Genesis-PickLift-Easy-autogen-1024
Episodes: 1024
Frames: 286444
FPS: 50 (control_dt=0.02 s)
Robot: Franka Panda with Robotiq 2F-85 gripper
Cameras: two 128×128 RGB wrist cameras
Language instruction:
Pick up the cube and lift it above the table
pick-cube-genesisThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "franka",
"total_episodes": 50,
"total_frames": 10000,
"total_tasks": 1,
"total_videos": 50,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 60,
"splits": {
"train": "0:50"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/jadechoghari/pick-cube-genesis.roboverse-pick_cube-genesis-E100This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"fps": 30,
"features": {
"observation.state": {
"dtype": "float32",
"shape": [
9
],
"names": [
"panda_finger_joint1",
"panda_finger_joint2",
"panda_joint1",
"panda_joint2",
"panda_joint3"… See the full description on the dataset page: https://huggingface.co/datasets/ases200q2/roboverse-pick_cube-genesis-E100.Genesis_AI_Code_50k
Genesis AI Code 50K (Expert)
Developed by: Within Us AI
Expert dataset with diff supervision, failure→reflection→correction, MoE labels, and governance flags.
Splits
train: 49,000
validation: 1,000
Highlights
Tests-as-truth supervision patterns
Diff-first patching
Agentic loops (plan→edit→test→reflect) with bounded budgets
Tool-call trace supervision (where present)
Governance/audit & policy-gate awareness
Storage format
Parquet… See the full description on the dataset page: https://huggingface.co/datasets/WithinUsAI/Genesis_AI_Code_50k.