datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
DAG-MATH-Formatted-CoT
Benchmark Overview
This dataset card contains 2,894 gold-standard DAG-MATH formatted CoT from problems from Omni-MATH.
Top‑Level Schema
Each JSON file is a list with a single object describing the problem:
problem_id: integer identifier of the problem.
domain: list of strings describing the topic taxonomy.
difficulty: numeric difficulty indicator from 1 (easiest) to 6 (hardest).
problem_text: problem statement.
sample_id: sample identifier for the solution trace.… See the full description on the dataset page: https://huggingface.co/datasets/yuanhezhang/DAG-MATH-Formatted-CoT.openvivqa-formating-vlm
OpenViVQA Formatting Dataset for VLM
A Vietnamese multimodal instruction-format dataset for training Vision Language Models (VLMs) on Visual Question Answering (VQA) tasks.
This dataset reformats OpenViVQA-style samples into conversational instruction-tuning format compatible with modern VLM training pipelines such as:
Qwen2-VL
LLaVA
InternVL
Phi-3 Vision
Idefics
SmolVLM
Dataset Structure
Each sample contains:
image: input image
conversations: multi-turn… See the full description on the dataset page: https://huggingface.co/datasets/Nhanvi282/openvivqa-formating-vlm.astrolabe-trace-format-test
Astrolabe trace format test (temporary)
numeric-format-catalog
Numeric Format Catalog (Trinity S3AI / t27)
Version: v3.0 (2026-06-13) -- supersedes v2 (count=81, withdrawn) and v1 (count=77, withdrawn).
Count: 83 numeric formats across 13 clusters.
License: CC-BY-4.0.
Maintainer: Trinity S3AI -- admin@t27.ai -- ORCID 0009-0008-4294-6159.
Companion dataset: playra/numeric-conformance-packs (bit-exact conformance vectors that instantiate this catalog).
Linked papers:
Catalog preprint (84-format ruler): arXiv:2606.09686 -- HF paper page… See the full description on the dataset page: https://huggingface.co/datasets/playra/numeric-format-catalog.Arabic_dataset_13M_translated_cleaned_v2_jsonl_format_ViT-B-16-SigLIP-512_validationexp-compounding-format
Compounding x Format: Specification Framing in Agentic Pipelines
Dataset Summary
Two experiments (1,440 LLM calls across 8 models) testing whether specification framing attenuates or amplifies dimensional collapse across multi-step agentic shopping pipelines. The dataset operationalises Section 5.16 ("Specification Paradox") of the companion R15 paper: Brand Function specification works in single-step contexts (reducing the Dimensional Collapse Index toward the… See the full description on the dataset page: https://huggingface.co/datasets/spectralbranding/exp-compounding-format.exp-bf-format
Brand Function Format Optimization (Exp D)
Dataset Summary
375 LLM responses (355 valid, 20 parse errors, 0 failures) testing which representational format of a brand function specification maximizes AI comprehension fidelity. Five formats (JSON structured, prose narrative, tabular minimal, ranked list, score-only vector) were crossed with five canonical SBT brands (Hermes, IKEA, Patagonia, Tesla, Erewhon), five model families (Claude Haiku 4.5, GPT-4o-mini… See the full description on the dataset page: https://huggingface.co/datasets/spectralbranding/exp-bf-format.Arabic_dataset_1M_translated_jsonl_format_ViT-B-16-plus-240This translation done using https://huggingface.co/Helsinki-NLP/opus-mt-en-ar
Arabic_dataset_13M_translated_cleaned_v2_jsonl_format_ViT-B-16-plus-240This dataset repo contains the dataset (CC3M+CC12M+SBU) translated using opus-mt-en-ar and cleaned. Its size about 13M
Arabic_dataset_13M_translated_cleaned_v2_jsonl_format_ViT-B-16-plus-240-fulldata-v2DatasetDict({
train: Dataset({
features: ['index', 'embeddings', 'en_caption', 'ar_caption', 'nr_words', 'url'],
num_rows: 12166802
})
})
ccs_synthetic_ar_1M-Arabic_dataset_1M_translated_jsonl_formatontocord__wide_3b_sft_stage1.2-ss1-expert_formatted_text-details
Dataset Card for Evaluation run of ontocord/wide_3b_sft_stage1.2-ss1-expert_formatted_text
Dataset automatically created during the evaluation run of model ontocord/wide_3b_sft_stage1.2-ss1-expert_formatted_text
The dataset is composed of 38 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/ontocord__wide_3b_sft_stage1.2-ss1-expert_formatted_text-details.ontocord__wide_3b_sft_stag1.2-lyrical_news_software_howto_formattedtext-merge-details
Dataset Card for Evaluation run of ontocord/wide_3b_sft_stag1.2-lyrical_news_software_howto_formattedtext-merge
Dataset automatically created during the evaluation run of model ontocord/wide_3b_sft_stag1.2-lyrical_news_software_howto_formattedtext-merge
The dataset is composed of 38 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/ontocord__wide_3b_sft_stag1.2-lyrical_news_software_howto_formattedtext-merge-details.openbmb__MiniCPM-S-1B-sft-llama-format-details
Dataset Card for Evaluation run of openbmb/MiniCPM-S-1B-sft-llama-format
Dataset automatically created during the evaluation run of model openbmb/MiniCPM-S-1B-sft-llama-format
The dataset is composed of 38 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/openbmb__MiniCPM-S-1B-sft-llama-format-details.