AI Training Datasets

Dataset · Robotics

tracker-pov

Eidon AI

1,274 hours of egocentric video paired with 7-point IMU arm tracking, recorded during ordinary household work. Contributors wore a head-mounted camera and a seven-sensor IMU harness while doing real chores in their own homes: laundry, cleaning, dishes, cooking. Each recording pairs first-person video with 24 Hz orientation data for both hands, both forearms, both upper arms, and the chest. This is a complete, final release. Eidon AI (Solidic Labs Inc) has wound down; the dataset is published under CC-BY-4.0 so it outlives the company. The release is split across two repos and a bucket. The data/NNN/ split exists only to keep directories manageable; the grouping carries no meaning. Both are…

Raw Inspect AI logs (.eval files) of every run made for runs. The intent is that no run can be left out afterwards. - Each file is at runs/ /.eval. - The manifest manifests/.json in the code repository records the sha256 of each file, the code version, the model files and the settings of the run, and the dataset commit at which the file was published and verified. - The files are published only after a scrub check: no key material, no home-directory paths, no private addresses or host names. Status: infrastructure in progress. There are no findings yet. The runs published so far are tests of the plumbing and measurements of the substrate, not results about monitoring. An.eval file is a zip…

Human gameplay from a first-person rooftop-courier game, recorded at the simulation tick rather than from the screen. One row per tick: the exact input the player gave, and the full world state that input produced. The action labels and the world state are ground truth read out of the simulator, not inferred from video. No pose estimation, no key-press detection, no frame differencing, no interpolation. injump is the jump bit the game consumed on that tick; vely is the number the physics step wrote. Every row is exact to the byte the recorder stored. Every session in this dataset has been replay-verified: re-simulated from its recorded seed and input stream and compared field-by-field…

The Math-Aware Retrieval benchmark from MathNet (paper): given an Olympiad problem, retrieve a mathematically equivalent reformulation of it from a large corpus of hard negatives. Each equivalent is provided at three difficulty tiers — easy, medium, hard — that increasingly disguise the surface form while preserving the underlying problem, so retrieval gets sharply harder even though the answer is the same. - 15,000 queries, one gold-equivalent target per tier. - 3 tiers (easy, medium, hard), over a shared 117,088-doc corpus (equivalents at all tiers + LLM near-miss hard negatives + distractors). - Standard BEIR / MTEB layout. All three configs share the same corpus and queries; they differ…

每条样本提供：RGB 源视频、XDoG 线稿视频（生成目标）、目标实例掩码（mask 版）、 SpaTracker 运动轨迹、首/末帧图、自然语言描述。 - 仓库文件：dataset1.zip（21.9 GB），解压后即本文档描述的 datasets/ 目录 （3 个 mask 版子集 + 3 个 dense 版子集，共 2,258 条索引记录） - 解压：unzip dataset1.zip -d SketchTraj/（解压后约 25 GB） 轨迹有两个版本，并存于 datasets/ 下： - mask 版（datasets/ /）：查询点限制在 SAM2 实例掩码内，聚焦主体运动 - dense 版（datasets/ dense/）：全图均匀网格 1024 点，覆盖整帧运动场 只做轨迹与可视化，视频 / 线稿 / 首末帧 / prompt 全部复用 mask 版（不复制文件， 索引里的 sourcevideo / gtvideo 仍指向 datasets/ /）： - trajvis/.mp4 是轨迹叠在 RGB 源视频上的可视化（H.264 / 832×480 / 49 帧 / 12fps）， trajvis/trajreport.json 给出每条的 visibleratio、meansteppx（mask 版还有 inmaskratio） 训练特征（preprocessed/：Wan2.1 VAE latent / UMT5-XXL / OpenCLIP / ycond）当前未产出， 数据集定稿后由 bash…

Dataset · Text retrieval

ipfs_ghana_laws_ir

Justicedao

Research retrieval release of endomorphosis/ipfsghanalaws (revision a1ce995b659ef704a91156115ea3aac1a714183b) packaged as country-laws-ir-graphrag/v1 (layout family skillcenter-huggingface-release/v3 / publicus-ir). Not legal advice. This is a research snapshot. The official gazette / authentic source of Ghana prevails over this corpus. Retrieved documents and graph edges are retrieval evidence only. No legal text was invented. Primary key: entrycid (CIDv1 raw sha2-256 of a canonical identity record). Integer documentindex values are compact shard pointers, not identities. Target Hub id (packaging metadata only): justicedao/ipfsghanalawsir. entrycid, lawcid, recordtype, jurisdiction…

Grid-Bench evaluates complete spatial state prediction. Each instance gives a language model an initial grid, explicit transition rules, and an ordered event history as text. The model predicts the complete final grid. Executable rules provide exact reference states, allowing both complete answers and individual cell values to be checked. Explore the project website and use the companion code repository to simulate, generate, Original data and documentation are licensed under CC BY 4.0. Collected model answers are not distributed; see licensing and release scope. Each configuration has equivalent Parquet and JSONL files. There are no training or validation splits. Instances cover PAINT…

Dataset · Text generation

csharp-ml-complation

Dvislobokov

Caret-based full-line completion samples extracted from permissively licensed C# repositories with csharp-dataset-prepare. Each sample is an exact editor position: leftcontext ends at the caret, targettext is the rest of the physical line (no newline), rightcontext follows it. Original source is reconstructable from offsets; every row carries repositoryid, revision, relativepath and license for attribution. Splits are assigned per repository group (forks/near-copies together); cross-repository duplicate files are kept once. No code was built or executed. See the builder repository for the data contract, leakage policy and benchmarks. For every sample the builder removes the hidden target…

Dataset

D14-codegen-annotated

AI4Manufacturing

[Placeholder] [Placeholder] [Placeholder] [Placeholder] [Placeholder] [Placeholder]

Dataset · Audio to audio

AURA-Chat-Edit

Open-Robust Lab

Overview AURA-Chat-Edit is a large-scale conversational music editing dataset used to train AURA, a unified multimodal framework for conversational music editing. The dataset contains 66,539 multi-turn dialogues pairing natural-language edit instructions with structured edit-token outputs across 7 edit types. Each dialogue simulates a user requesting a music edit (e.g., "remove the drums", "add a jazzy

A preference dataset for THREATINTELLIGENCE, harvested from real, human-labelled sources and curated by an automated harvesting harness with an LLM quality gate. Standard preference / DPO schema — each row: 80/10/10 train / validation / test (seeded shuffle): train:1632 / validation:204 / test:204 - synthetic:deepseek/deepseek-v4-flash-0731 prompt and chosen are real — harvested from the human-labelled sources listed above (accepted / upvoted / high-rated responses). rejected is partly synthetic. Most sources retain only the accepted answer, so where no genuine low-scored counterpart existed the negative was generated by defect injection: an LLM rewrites that row's own chosen with exactly…

A preference dataset for SANCTIONSSCREENINGMATCH, harvested from real, human-labelled sources and curated by an automated harvesting harness with an LLM quality gate. Standard preference / DPO schema — each row: 80/10/10 train / validation / test (seeded shuffle): train:1553 / validation:194 / test:195 - https://law.stackexchange.com/questions/103190 - https://law.stackexchange.com/questions/103256 - https://law.stackexchange.com/questions/104418 - https://law.stackexchange.com/questions/104851 - https://law.stackexchange.com/questions/105647 - https://law.stackexchange.com/questions/105745 - https://law.stackexchange.com/questions/106488 - https://law.stackexchange.com/questions/106690…