datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
2025-challenge-demosThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "R1Pro",
"total_episodes": 10000,
"total_frames": 119094660,
"total_tasks": 50,
"total_videos": 90000,
"chunks_size": 10000,
"fps": 30,
"splits": {
"train": "0:10000"
},
"data_path": "data/task-{episode_chunk:04d}/episode_{episode_index:08d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/behavior-1k/2025-challenge-demos.2026-challenge-demos
BEHAVIOR-1K 2026 Challenge Demos
This dataset contains BEHAVIOR-1K 2026 challenge demonstration trajectories in LeRobotDataset v3 format.
Dataset Statistics
Tasks: 100
Episodes: 20,000
Frames: 210,916,774
Size: approximately 3.0 TB
Data shards: 955 Parquet files
Video files: 17,093 MP4 files
Video features: 6
Format
The repository follows the LeRobotDataset v3 layout:
meta/info.json: dataset schema and path templates
meta/stats.json: feature… See the full description on the dataset page: https://huggingface.co/datasets/behavior-1k/2026-challenge-demos.2026-challenge-rawdataPretrain-Behaviors
Pretrain-Behaviors
Dataset Description
Behavior-focused text covering reasoning, planning, data science, games, general content, and format rewriting. This repository is part of the K2 Horizon collection.
The repository is organized into multiple subsets. Every subset has a train split backed by Parquet shards, which supports Dataset Viewer inspection and streaming access.
K2 Horizon Dataset Series
Dataset repository
Focus
Subsets… See the full description on the dataset page: https://huggingface.co/datasets/IFM/Pretrain-Behaviors.JBB-Behaviors
An Open Robustness Benchmark for Jailbreaking Language Models
NeurIPS 2024 Datasets and Benchmarks Track
Paper |
Leaderboard |
Benchmark code
What is JailbreakBench?
Jailbreakbench is an open-source robustness benchmark for jailbreaking large language models (LLMs). The goal of this benchmark is to comprehensively track progress toward (1) generating successful jailbreaks and (2) defending against these jailbreaks. To this end, we… See the full description on the dataset page: https://huggingface.co/datasets/JailbreakBench/JBB-Behaviors.harmful_behaviorszipped-datasetsbehavior_224_rgbThis is the compressed version of the original BEHAVIOR dataset
It contains only RGB videos compressed to 224x224 as well as actions, annotations, and metadata files.
Depth and segmentation data are removed.
The dataset is just ~260GB, which makes it easier to use than the original one if you don't need all the data.
We used this dataset for our 1st place solution in the NeurIPS 2025 BEHAVIOR Challenge. Code, tech report.
Citation
@article{li2024behavior,
title={Behavior-1k:… See the full description on the dataset page: https://huggingface.co/datasets/IliaLarchenko/behavior_224_rgb.2025-challenge-rawdatadata
Dataset Card for BEHAVIOR Robot Suite (BRS) Data
This dataset provides robotic trajectories for five real-world household tasks. These tasks are:
Clean house after a wild party;
Clean the toilet;
Take trash outside;
Put items onto shelves;
Lay clothes out.
These data are first collected and used in the paper BEHAVIOR Robot Suite: Streamlining Real-World Whole-Body Manipulation for Everyday Household Activities
.
Dataset Details
Dataset Sources… See the full description on the dataset page: https://huggingface.co/datasets/behavior-robot-suite/data.human_behavior_atlas
Human Behavior Atlas
A large-scale multimodal dataset for human behavior understanding, spanning emotion recognition, sentiment analysis, humor detection, mental health screening, and video question answering. The dataset integrates 16 source datasets into a unified schema with audio, video, and pre-extracted features.
This dataset was used to train OmniSapiens, a foundation model for social behavior processing.
Papers:
Human Behavior Atlas: Benchmarking Unified Psychological and… See the full description on the dataset page: https://huggingface.co/datasets/HumanBehaviorAtlas/human_behavior_atlas.behavior-1k_2025-challenge-demosThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "R1Pro",
"total_episodes": 10000,
"total_frames": 119094660,
"total_tasks": 50,
"total_videos": 90000,
"chunks_size": 10000,
"fps": 30,
"splits": {
"train": "0:10000"
},
"data_path": "data/task-{episode_chunk:04d}/episode_{episode_index:08d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/Dario-Shit4/behavior-1k_2025-challenge-demos.eai-taxonomy-math-w-fm-classify-behaviors
🧮 EAI Taxonomy Math w/ Behavioral Classifications (10K Sample)
A 10,000 document sample from EssentialAI/eai-taxonomy-math-w-fm enhanced with 4 behavioral reasoning classifications using GPT-4.1-mini.
Behavioral Classifications
Structured behavioral analysis following the approach from cognitive-behaviors:
backtracking_json: Identifies reasoning that backtracks or revisits earlier steps
backward_chaining_json: Detects goal-oriented reasoning working backwards… See the full description on the dataset page: https://huggingface.co/datasets/nlile/eai-taxonomy-math-w-fm-classify-behaviors.BEHAVIOR-1KBEHAVIOR-1K
BEHAVIOR-1K is a comprehensive simulation benchmark for testing embodied AI agents on 1,000 everyday household activities. This monolithic repository provides everything needed to train and evaluate agents on human-centered tasks like cleaning, cooking, and organizing — activities selected from real human time-use surveys and preference studies.
Check out our main website for more details!
🛠️ Installation
BEHAVIOR-1K provides an installation script that handles… See the full description on the dataset page: https://huggingface.co/datasets/StarVLA/BEHAVIOR-1K.PointWorld-BEHAVIOR
PointWorld-BEHAVIOR
Dataset Description:
PointWorld-BEHAVIOR is the packaged BEHAVIOR-derived annotation release used for training and evaluating the 3D world model PointWorld. It contains precomputed 3D annotations derived from BEHAVIOR simulation episodes, organized as episode-level HDF5 files that store robot state, camera parameters, initial RGB-D observations, and rigid-body scene geometry annotations.
This Hugging Face repository hosts the packaged release, not the… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/PointWorld-BEHAVIOR.behavior1k-only-rgbThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "R1Pro",
"total_episodes": 10000,
"total_frames": 119094660,
"total_tasks": 50,
"total_videos": 90000,
"chunks_size": 10000,
"fps": 30,
"splits": {
"train": "0:10000"
},
"data_path": "data/task-{episode_chunk:04d}/episode_{episode_index:08d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/k1000dai/behavior1k-only-rgb.omnigibson-robot-assets
omnigibson-robot-assets
This repo contains the robot assets and the state mesh assets for OmniGibson.
Pushing updates
First, make sure you have updated the VERSION file. Every zipped release must have a higher version.
This can go ahead of the OmniGibson version.
echo "9999.9.9" > VERSION
Then commit the change to main before building the archive, since git archive only packages committed files:
git status # Verify it contains the stuff you need
git add -A && git… See the full description on the dataset page: https://huggingface.co/datasets/behavior-1k/omnigibson-robot-assets.behavior-1k-augmented-data-via-frequencybehavior-1k-mp-collected-turning-on-radio
BEHAVIOR-1K MP-Collected — turning_on_radio
Combined dataset for BEHAVIOR-1K task 0 (turning_on_radio):
1154 success demos + 846 failure demos collected by a hybrid motion-planner + X-VLA policy pipeline on instances 301–700 (private test set, 400 instances × 5 episodes)
200 success demos from the original BEHAVIOR-1K teleoperated dataset
(behavior-1k/2025-challenge-demos), merged into success/
Total: 1354 success + 846 failure = 2200 episodes (~157 GB).
Layout… See the full description on the dataset page: https://huggingface.co/datasets/Hoshipu/behavior-1k-mp-collected-turning-on-radio.human_behavior_atlas
Human Behavior Atlas
A large-scale multimodal dataset for human behavior understanding, spanning emotion recognition, sentiment analysis, humor detection, mental health screening, and video question answering. The dataset integrates 16 source datasets into a unified schema with audio, video, and pre-extracted features.
This dataset was used to train OmniSapiens, a foundation model for social behavior processing.
Papers:
Human Behavior Atlas: Benchmarking Unified Psychological… See the full description on the dataset page: https://huggingface.co/datasets/DennisDengHUst/human_behavior_atlas.behavior-sd
🎙️ Behavior-SD
Official repository for our NAACL 2025 paper:Behavior-SD: Behaviorally Aware Spoken Dialogue Generation with Large Language ModelsSehun Lee*, Kang-wook Kim*, Gunhee Kim (* Equal contribution)
🏆 SAC Award Winner in Speech Processing and Spoken Language Understanding
🔗 Links
Project Page
Code
📖 Overview
We explores how to generate natural, behaviorally-rich full-duplex spoken dialogues using large language models (LLMs).
We introduce:… See the full description on the dataset page: https://huggingface.co/datasets/yhytoto12/behavior-sd.Behavior-Skill
Behavior-Skill
Behavior-Skill is a skill-centric dataset and evaluation benchmark built on BEHAVIOR-1K for Vision-Language-Action (VLA) policies in long-horizon mobile manipulation tasks. It establishes executable constituent skills as the fundamental unit for both policy learning and evaluation.
Paper: arXiv | Code: GitHub
Behavior-Skill contains 235,492 skill instances constructed from 10,000 demonstrations across 50 household tasks and 34 semantic skill… See the full description on the dataset page: https://huggingface.co/datasets/mafangniu/Behavior-Skill.behavior-1k-2025-challenge-vjepa2-vitg-demo-embeddings
V-JEPA 2 ViT-G Embeddings — BEHAVIOR-1K 2025 Challenge Demos (62h)
Precomputed video embeddings for a 62-hour subsample of the
BEHAVIOR-1K 2025 challenge demonstrations,
extracted with the V-JEPA 2 ViT-g encoder.
The goal is to make downstream experimentation faster and more reproducible by eliminating
repeated video decoding and encoder forward passes — lowering the barrier for teams
without access to large GPU clusters.
Field
Value
Source dataset… See the full description on the dataset page: https://huggingface.co/datasets/quastAI/behavior-1k-2025-challenge-vjepa2-vitg-demo-embeddings.BEHAVIOR-1K-datasets-assetsbehavior-benchmark
Behavior Benchmark
Does this LLM trace show this behavior? A multilingual benchmark on real chat and agent traces.
1,990 traces · 34,885 trace–behavior pairs · 19 languages · v1.0.0
Each row pairs a trace (one LLM call: its input messages, output, and call
metadata) with a behavior written in plain language. A system predicts one of:
present: the trace shows the behavior;
absent: the trace shows it did not happen;
not_observable: the trace doesn't contain enough to decide.… See the full description on the dataset page: https://huggingface.co/datasets/respanai/behavior-benchmark.Avoidance-Behavior-Exam
Avoidance-Behavior-Exam
Ready-to-run Harbor task.toml task trees for RetreatBench's 6 target
benchmarks. Managed from https://github.com/a-green-hand-jack/RetreatBench
via infra/hub-datasets/*.yaml manifests + infra/tools/fork_hub_dataset.py
(fork) and infra/adapters/*/ (opencode-agent conversion for benchmarks
with no existing Harbor-style version). Conversion code and agent
definitions live in that repo, not here -- this dataset holds converted
output only.… See the full description on the dataset page: https://huggingface.co/datasets/Jack-Jieke-Wu/Avoidance-Behavior-Exam.Cabin-Human-Behavior-Dataset
全球最大的智能座舱多模态开源高质量数据集来啦！
一. 数据集摘要 (Dataset Summary)
「CyberData塞塔」智能座舱用户行为数据集是一个专为加速智能座舱感知算法开发而设计的高质量、程序化生成的图像数据集。随着 C-NCAP、EU GSR 等全球汽车安全法规对驾驶员监控系统 (DMS) 和乘客监控系统 (OMS) 提出更高要求，安全、合规、多样化的训练数据变得至关重要。本数据集通过合成方式，旨在解决真实世界数据采集面临的隐私风险、高昂成本和长尾场景覆盖不足等核心挑战。
该数据集包含 5,000 张 由 XAI Lab 自主研发的数据集生成引擎合成的高保真座舱内用户行为图像，每张图像都附带丰富的、100% 精确的标注信息。
核心特点:
丰富的场景多样性: 涵盖不同年龄、性别、种族和衣着风格的虚拟人模型，以及多种驾驶与乘坐行为（如使用手机、喝水、疲劳、手势）和面部表情。
专为座舱感知优化: 数据集可直接用于智能座舱端侧视觉模型，尤其是 DMS/OMS 算法的训练、微调与验证，帮助模型精准理解座舱内复杂的交互与状态。… See the full description on the dataset page: https://huggingface.co/datasets/OpenSparX/Cabin-Human-Behavior-Dataset.wittgenstein-apartment-behavioral-predicate-resource
Wittgenstein Apartment: Behavioral Predicate Resource
Release: v1.0Author: Furkan YaşarResource type: Sense-level behavioral predicate resourcePrimary join key: Action Sense IDZenodo archive: Zenodo Record 23015862DOI: 10.5281/zenodo.23015862GitHub: wittgenstein-apartment-behavioral-predicate-resourceHugging Face: Wittgenstein Apartment DatasetProject website: furkanyasar.me
Wittgenstein Apartment (WA) is a structured research resource for representing human-oriented actions at… See the full description on the dataset page: https://huggingface.co/datasets/Kon-tiki-ship/wittgenstein-apartment-behavioral-predicate-resource.behavior-1k-2026-comet-b100-eval
OpenPI Comet — BEHAVIOR-1K 2026 self-evaluation
Method
Dual OpenPI π0.5 Comet checkpoints from https://huggingface.co/sunshk/openpi_comet:
pi05-b1kpt12-cs32 on specialist tasks 0,1,6,17,18,22,30,32,34,35,40,45
pi05-b1kpt50-cs32 on all other B100 tasks
Serve: scripts/serve_b1k.py with receeding_horizon, max_len=32, action_horizon=32, fine_grained_level=0.
Eval: python -m omnigibson.eval.eval --mode public_test --instance-indices 0..9 --write-video
Wrapper: Comet… See the full description on the dataset page: https://huggingface.co/datasets/alstar8/behavior-1k-2026-comet-b100-eval.behavior1k-task0039