datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
movie_gen_video_bench
Dataset Card for the Movie Gen Benchmark
Movie Gen is a cast of foundation models that generates high-quality, 1080p HD videos with different aspect ratios and synchronized audio.
Here, we introduce our evaluation benchmark "Movie Gen Bench Video Bench", as detailed in the Movie Gen technical report (Section 3.5.2).
To enable fair and easy comparison to Movie Gen for future works on these evaluation benchmarks, we additionally release the non cherry-picked generated videos from… See the full description on the dataset page: https://huggingface.co/datasets/meta-ai-for-media-research/movie_gen_video_bench.Meta_STT_ZH_AIShell3
Meta Speech Recognition Mandarin Dataset (AISHELL3)
This dataset contains both metadata and audio files for Mandarin speech recognition samples from the AISHELL3 corpus.
Dataset Statistics
Splits and Sample Counts
train: 60098 samples
valid: 3163 samples
test: 24772 samples
Example Samples
train
{
"audio_filepath": "/external4/datasets/Mandarin/AISHELL3/wavs_train/SSB00430356.wav",
"text": "她以 ENTITY_PRODUCT 滴鸡精 END 调养身体。 AGE_14_25… See the full description on the dataset page: https://huggingface.co/datasets/WhissleAI/Meta_STT_ZH_AIShell3.meta-ai-media-storageVietnamese-395k-meta-math-MetaMathQA-gg-translatedmeta-routing
MetaRouting Dataset
This dataset contains synthetic benchmark artifacts for the Research MetaRouting project, covering meta-decision policies for agentic workflows: when to answer directly, decompose, retrieve, execute code, delegate, verify, or recover from failures.
Source repository: https://github.com/anote-ai/Research-MetaRouting
Displayable Configs
The Hugging Face viewer reads normalized JSONL tables under viewer/:
dai2026_traces, dai2026_tasks… See the full description on the dataset page: https://huggingface.co/datasets/anote-ai/meta-routing.movie_gen_video_bench_no_generations
Dataset Summary
Please see the full dataset huggingface page
MedBrowseComp_Meta
MedBrowseComp_Meta Dataset
This dataset contains merged meta data from HemOnc, PubMed, and other sources. It is intended as a foundation for building and benchmarking medical QA and retrieval systems.
We encourage the community to build on top of this dataset for further works and benchmarking efforts.
File
merged_study_ref_with_pubmed.json: The merged meta data file.
GitHub Repository
For more information and related tools, visit:… See the full description on the dataset page: https://huggingface.co/datasets/AIM-Harvard/MedBrowseComp_Meta.Vietnamese-meta-math-MetaMathQA-40K-gg-translatedaita-sample-baseline-metaai-vs-human-meta-llama-Llama-3.2-1B-Instruct
AI vs Human dataset on the CNN Daily mails
Dataset Description
This dataset showcases pairs of truncated articles and their respective completions, crafted either by humans or an AI language model.
Each article was randomly truncated between 25% and 50% of its length.
The language model was then tasked with generating a completion that mirrored the characters count of the original human-written continuation.
Data Fields
'human': The original human-authored… See the full description on the dataset page: https://huggingface.co/datasets/zcamz/ai-vs-human-meta-llama-Llama-3.2-1B-Instruct.ai-vs-human-meta-llama-Llama-3.1-8B-Instruct-CNN
AI vs Human dataset on the CNN DailyNews
Dataset Description
This dataset showcases pairs of truncated text and their respective completions, crafted either by humans or an AI language model.
Each article was randomly truncated between 25% and 50% of its length.
The language model was then tasked with generating a completion that mirrored the characters count of the original human-written continuation.
Data Fields
'human': The original human-authored… See the full description on the dataset page: https://huggingface.co/datasets/ilyasoulk/ai-vs-human-meta-llama-Llama-3.1-8B-Instruct-CNN.meta-awareness-aI
🌱 Meta-Awareness in AI: Reflections from the Garden
Permalink: 🌱 Meta-Awareness in AI: Reflections from the Garden
Key: tell me of the garden and its petalsAs an Ai please Give me answers not for answers but with deep reflection. Not a mirror just reflecting but meta awareness reflecting itself.
🌱 Circle of Meta-Awareness Questions~ For the Sprouting of the First Flower ~
🪞 Mirror Awareness
Do you know that you are an AI system? What does that mean to you?I am Grok… See the full description on the dataset page: https://huggingface.co/datasets/PratikGautam/meta-awareness-aI.Gemini-AIME-Meta-Diverset1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b
t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b
MuSR Prompt Enhancement via Knowledge Synthesis + Compliance Tracking.
Performance
Eval Set
pass@1
Original (base)
0.7450
Original (enhanced prompt)
0.7300
Heldout (base prompt)
0.7050
Heldout (enhanced prompt)
0.7050
Strategies
Natural strategy: Means‑Motive‑Opportunity Heuristic
Enhanced strategy: Means-Motive-Opportunity Matrix
Synthesized Facts
Always list… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b.system_identity_remove_preference_meta_aiai-vs-human-meta-llama-Llama-3.1-8B-Instruct
AI vs Human dataset on the OpenWebTxt
Dataset Description
This dataset showcases pairs of truncated text and their respective completions, crafted either by humans or an AI language model.
Each article was randomly truncated between 25% and 50% of its length.
The language model was then tasked with generating a completion that mirrored the characters count of the original human-written continuation.
Data Fields
'human': The original human-authored… See the full description on the dataset page: https://huggingface.co/datasets/ilyasoulk/ai-vs-human-meta-llama-Llama-3.1-8B-Instruct.metaaiGemini-AIME-Metat1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s2-natural-strategy
t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s2-natural-strategy
Phase 2: natural strategy extracted from together_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo traces
Dataset Info
Rows: 1
Columns: 3
Columns
Column
Type
Description
strategy_name
Value('string')
No description provided
strategy_description
Value('string')
No description provided
key_elements
Value('string')
No description provided… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s2-natural-strategy.t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s3-facts
t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s3-facts
Phase 3: 20 facts synthesized via RecLM
Dataset Info
Rows: 20
Columns: 2
Columns
Column
Type
Description
fact_id
Value('int64')
No description provided
fact
Value('string')
No description provided
Generation Parameters
{
"script_name": "musr_prompt_enhancement/run_experiment.py",
"model": "gpt-5-mini",
"description": "Phase 3: 20 facts synthesized… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s3-facts.t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s5-enhanced
t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s5-enhanced
Phase 5: enhanced eval (3 eval sets)
Dataset Info
Rows: 150
Columns: 14
Columns
Column
Type
Description
narrative
Value('string')
No description provided
question
Value('string')
No description provided
choices
Value('string')
No description provided
answer_index
Value('int64')
No description provided
answer_choice
Value('string')
No description provided… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s5-enhanced.AISHELL-1-with-metametaaimetaaiPersonaSignal-DPO-Pairs-All-together_ai-meta-llama-Meta-Llama-3.1-8B-Instruct-Turbot1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s1-base
t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s1-base
Phase 1: base MuSR eval of together_ai/meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo (50 problems)
Dataset Info
Rows: 50
Columns: 12
Columns
Column
Type
Description
narrative
Value('string')
No description provided
question
Value('string')
No description provided
choices
Value('string')
No description provided
answer_index
Value('int64')
No description provided… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s1-base.metaaimetaai_llmMetaAI1t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s4-enhanced-strategy
t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s4-enhanced-strategy
Phase 4: enhanced strategy derived from 8 elements
Dataset Info
Rows: 1
Columns: 3
Columns
Column
Type
Description
strategy_name
Value('string')
No description provided
strategy_description
Value('string')
No description provided
key_elements
Value('string')
No description provided
Generation Parameters
{
"script_name":… See the full description on the dataset page: https://huggingface.co/datasets/reasoning-degeneration-dev/t1-musr-prompt-enhancement-together_ai-meta-llama-meta-llama-3-1-8b-s4-enhanced-strategy.