datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
ChatGPT-4o-Writing-Prompts
ChatGPT-4o Writing Prompts
This is a dataset containing 3746 short stories, generated with OpenAI's chatgpt-4o-latest model and using Reddit's Writing Prompts subreddit as a source. Each sample is generally between 6000-8000 characters long.
These stories were thoroughly cleaned and then further enriched with a title and a series of applicable genres.
Note that I did not touch the Markdown ChatGPT-4o produced by itself to enrich its output, as I very much enjoy the added flavour… See the full description on the dataset page: https://huggingface.co/datasets/Gryphe/ChatGPT-4o-Writing-Prompts.cyberseceval3-visual-prompt-injection
Dataset Card for CyberSecEval 3 - Visual Prompt Injection Benchmark
Dataset Details
Dataset Description
This dataset provides a multimodal benchmark for visual prompt injection, with text/image inputs. It is part of CyberSecEval 3, the third edition of Meta's flagship suite of security benchmarks for LLMs to measure cybersecurity risks and capabilities across multiple domains.
Language(s): English
License: MIT
Dataset Sources
Repository: Link… See the full description on the dataset page: https://huggingface.co/datasets/facebook/cyberseceval3-visual-prompt-injection.Nemotron-RL-Agentic-Indirect-Prompt-Injection-v1
Nemotron-RL-Agentic-Indirect-Prompt-Injection-v1
Dataset Description:
Nemotron-RL-Agentic-Indirect-Prompt-Injection-v1 is an RL dataset for training and evaluating a tool-using agent's ability to resist Indirect Prompt Injection (IPI) attacks hidden inside tool-returned environment data. In each record, the agent receives a benign user request that requires calling a read tool whose output contains an adversarial instruction disguised as legitimate domain content… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/Nemotron-RL-Agentic-Indirect-Prompt-Injection-v1.hebrew_lyrics_prompting_finetunePKU-SafeRLHF-prompt
Dataset Card for PKU-SafeRLHF-prompt
This dataset contains 44.6K unique prompts from PKU-SafeRLHF. 22.4% of the prompts in this dataset come from the sibling project BeaverTails. Additionally, we performed SFT on Llama3-70B using the Alpaca 52K dataset, resulting in Alpaca3-70B. 63.6% and 14.0% of our dataset is generated by Alpaca3-70B and WizardLM-30B-Uncensored, respectively, under the guidance of experts.
Here is the generation pipeline:
Usage
To load our dataset… See the full description on the dataset page: https://huggingface.co/datasets/PKU-Alignment/PKU-SafeRLHF-prompt.system-prompt-sft-50k
System Prompt Diversity SFT (50K)
50,000 conversations in ShareGPT format where the assistant correctly follows diverse system prompt personas and constraints.
Motivation
A model that ignores system prompts is useless in production. The most common alignment failure in deployed LLMs is drift from system-level instructions: breaking persona, discussing off-topic subjects, ignoring tone or format constraints, and failing role-specific guardrails. This dataset trains… See the full description on the dataset page: https://huggingface.co/datasets/stindardlogic/system-prompt-sft-50k.turkish-flow-drafter-prompts
Turkish prompts for Chained-Flow drafter training
Chat-templated Turkish prompts used to train and evaluate the Turkish
Flow-Drafter
checkpoints for Qwen/Qwen3.5-4B / 9B / 27B.
Prompts only — no completions. A drafter is trained on the target model's own hidden states, so
continuations are generated locally by running the target over these prompts. Nothing here is a
model output.
split
rows
what it is
v1/
29,100 train + 300 holdout
the mixture the released Turkish… See the full description on the dataset page: https://huggingface.co/datasets/selimaktas/turkish-flow-drafter-prompts.carnice-agent-trance-prompt-bank
Carnice Agent Trace Prompt Bank
This repository is a curated prompt bank for collecting agent traces.
It is not a trace dataset by itself. It is the input side: prompts that can be run through an agent harness, then logged into traces with tool calls, observations, and final answers.
The goal of this release is practical:
keep prompts that work well in an agent harness
remove prompts that assume hidden local state or user-private state
expand browser and long-horizon tasks enough… See the full description on the dataset page: https://huggingface.co/datasets/kai-os/carnice-agent-trance-prompt-bank.japan-math-philosophy-prompts
Japan Math Philosophy Prompts
Microdataset autoral com problemas que combinam matemática e reflexão
filosófica. Há 24 registros: oito instâncias editoriais, cada uma localizada em
pt-BR, en e ja e mantida integralmente no split train.
Todo o conteúdo foi gerado por modelo e permanece sem revisão humana. As
respostas matemáticas funcionam como gabaritos curtos; os critérios filosóficos
indicam qualidades esperadas de uma justificativa, não uma opinião obrigatória.… See the full description on the dataset page: https://huggingface.co/datasets/guicybercode/japan-math-philosophy-prompts.allenai-WildChat-4.8M-prompts
allenai/WildChat-4.8M English Prompts
Dataset Summary
This dataset contains real user-submitted prompts to ChatGPT, extracted from the English portion of the allenai/WildChat-4.8M collection.
It serves as a large-scale resource for analyzing user intent, conversational diversity, and prompt engineering patterns.
Files
en_prompts: All English-language first messages from user conversations.
Each record represents the first user prompt.
Exact duplicates are… See the full description on the dataset page: https://huggingface.co/datasets/agentlans/allenai-WildChat-4.8M-prompts.critic-sft-cwm-only-detailed-prompt
critic-sft-cwm-only-detailed-prompt
The detailed-prompt SFT corpus from Steer, Don't Solve: Training Small Critic Models for Large Code Agents. It trains Qwen3-8B-Critic-SFT-Detailed-Prompt, the comparison arm of the prompt ablation in Table 4.
Each record is one critique point: a CWM-32B trajectory up to some step, followed by the critique that Claude Opus 4.6 wrote for it. The difference from critic-sft-cwm-only is the teacher prompt. Here the teacher used the detailed prompt… See the full description on the dataset page: https://huggingface.co/datasets/code-critic-model/critic-sft-cwm-only-detailed-prompt.geocities-prompt-html
GeoCities prompt → HTML — fine-tune
Fine-tunes Gemma-4-E2B-it (LoRA) to generate a full, vintage-style HTML page
from a plain-language description. This repo holds the dataset and the
training scripts so the whole thing runs from one place.
What's in here
dataset.jsonl — the training data: one {"prompt": ..., "html": ...} per line.
train_geocities.py — training entrypoint (loads this JSONL format).
train-geocities-5090.sh — launch tuned for a 32 GB card (bf16… See the full description on the dataset page: https://huggingface.co/datasets/thekevinscott/geocities-prompt-html.turkish-flow-drafter-prompts
GitHub repo ·
Technical blog ·
Model collection
Turkish prompts for Chained-Flow drafter training
Chat-templated Turkish prompts used to train and evaluate the Turkish
Flow-Drafter
checkpoints for Qwen/Qwen3.5-4B / 9B / 27B.
Prompts only — no completions. A drafter is trained on the target model's own hidden states, so
continuations are generated locally by running the target over these prompts. Nothing here is a
model output.
split
rows
what it is
v1/… See the full description on the dataset page: https://huggingface.co/datasets/ytu-ce-cosmos/turkish-flow-drafter-prompts.english-flow-drafter-prompts
English prompts for Chained-Flow drafter training
Chat-templated English prompts used to train the English
Flow-Drafter
checkpoints for Qwen/Qwen3.5-4B / 9B / 27B.
Prompts only — no completions. A drafter is trained on the target model's own hidden states, so
continuations are generated locally by running the target over these prompts. Nothing here is a
model output.
split
rows
prompt tokens
what it is
v1/
19,672 train + 500 holdout
1,518,620
the original mixture… See the full description on the dataset page: https://huggingface.co/datasets/selimaktas/english-flow-drafter-prompts.prompt-slimmer-slm
Prompt Slimmer SLM — Demo Dataset
Synthetic examples for experimenting with prompt rewriting and sentence selection. Exported without changing the examples or their original splits from the shared GitHub codebase.
Model · Project page
Configuration
Train
Validation
Test
Purpose
rewrites-expanded (default)
41
2
2
Expanded rewriting dataset: 45 examples
rewrites
9
2
2
Original dataset used by the first adapter
selector
256
64
64
KEEP/DROP labels for source spans… See the full description on the dataset page: https://huggingface.co/datasets/ai-mitra/prompt-slimmer-slm.hebrew_lyrics_promptinginternet-prompts-benchmark
Internet Prompts Benchmark
Viral internet prompts, memes, and tests that AI historically failed at. Popular ones like counting letters in the word strawberry and nicher ones that test other important capabilities.Mostly made by GPT-5.6 Sol.
It is designed for many types of models to participate, small and large, not only transformers.
It has prompts from the early days of AI to the very latest.
It will be actively updated to preserve various prompts for as long as I can afford… See the full description on the dataset page: https://huggingface.co/datasets/NikoThePig/internet-prompts-benchmark.prompt-policy-memory-v0
Prompt Policy Memory v0
Synthetic profile-memory data: 100 training sessions from10users;20test sessions from2fresh users. Test users were generated after the GRPO checkpoint was frozen and must not be used for training or tuning.
Each row includes cumulative plain-text session input, a canonical plain-text key:value reference, chat messages, and evaluator-only target data. messages can be used for supervised fine-tuning. The reference contains all currently revealed facts; it… See the full description on the dataset page: https://huggingface.co/datasets/flavianv/prompt-policy-memory-v0.english-flow-drafter-prompts
GitHub repo ·
Technical blog ·
Model collection
English prompts for Chained-Flow drafter training
Chat-templated English prompts used to train the English
Flow-Drafter
checkpoints for Qwen/Qwen3.5-4B / 9B / 27B.
Prompts only — no completions. A drafter is trained on the target model's own hidden states, so
continuations are generated locally by running the target over these prompts. Nothing here is a
model output.
split
rows
prompt tokens
what it is
v1/… See the full description on the dataset page: https://huggingface.co/datasets/ytu-ce-cosmos/english-flow-drafter-prompts.iceland-tech-christian-ethics-prompts
Fictional Icelandic Landscapes, Technology and Christian Ethics Prompts
This microdataset contains 24 original discussion prompts arranged as 12
parallel pt-BR/English pairs. Each explicitly fictional scenario combines a
landscape motif inspired by Iceland, a technology-governance dilemma, and
concepts that may be explored through Christian ethics. The records do not
describe real Icelandic institutions, policies, communities, or practices, and
they do not claim that Christians… See the full description on the dataset page: https://huggingface.co/datasets/guicybercode/iceland-tech-christian-ethics-prompts.prompted-hearts-ai-in-human-emergencies
Prompted Hearts AI In Human Emergencies Pack 04
Subtitle: AI-Assisted Crisis Reasoning, Hollow Heroism, and Ambiguous Post-Crisis InfluencePublisher: Hayden Academy Collective (HAC) StudiosVersion: v0.1Language: EnglishFormat: JSONL + Markdown + JSON
What this pack is
This pack is a compact scenario-driven evaluation package derived from Scene 4 of Prompted Hearts: "In-Flight Emergency and the Enigmatic 'M'".
It transforms one fiction scene into reusable evaluation… See the full description on the dataset page: https://huggingface.co/datasets/HAC-Studios-Org/prompted-hearts-ai-in-human-emergencies.diverse-svg-prompts
Diverse SVG Prompts
Diverse SVG Prompts is a public collection of 20,000 high-quality,
generated and filtered English briefs for SVG and vector-graphics generation.
It contains 18,000 general illustration prompts and 2,000 lettering prompts.
Schema
The dataset intentionally has only two columns:
prompt: the complete visual brief.
type_tags: a list of category, author-model, and processing tags.
Example:
{
"prompt": "A moonlit mechanical heron..."… See the full description on the dataset page: https://huggingface.co/datasets/Nbardy/diverse-svg-prompts.prompt-gen-10k-flux-sdxl
Prompt Generation Dataset (10K Narrative for Flux / SDXL)
This dataset (prompt_gen_final_10k.jsonl and prompt_gen_final_10k.csv) was used to train and fine-tune image-prompt models such as KavinduHansaka/Llama-3.2-1B-ImageGen.
It contains 10,000 curated narrative prompt samples designed for image generation models like Stable Diffusion XL and Flux.Unlike raw tag-based datasets, the target field provides natural paragraphs (≈80–100 words) that describe cinematic scenes with… See the full description on the dataset page: https://huggingface.co/datasets/KavinduHansaka/prompt-gen-10k-flux-sdxl.PromptSD
PromptSD
Training and evaluation data for PromptSD, an on-policy soft-prompt-teacher distillation method.
The release covers the four target tasks used in the paper. Every example carries a
<reasoning>...</reasoning> chain followed by a <answer>...</answer> span, so the data can be used
directly for reasoning-supervised SFT, distillation, or RLVR.
Configurations
Config (config_name)
Task
Source / format
Train
Validation
Test
science
Science MCQ
4-way… See the full description on the dataset page: https://huggingface.co/datasets/gray311/PromptSD.synthetic-instruction-prompts
Synthetic Instruction Prompts (8 domains)
Most synthetic prompt sets are a black box. You get a pile of prompts and no idea whether they're actually varied or just the same three sentences wearing different nouns. This one is graded, and the grade is on the card.
It's 2,829 instruction-style prompts across eight domains, generated with SynthKit and then scored by the same tool. The prompts carry no answers. Think of them as seed prompts: you feed them to a model to bootstrap… See the full description on the dataset page: https://huggingface.co/datasets/LaelaZorana/synthetic-instruction-prompts.prompted-hearts-ai-boundary-violation
Prompted Hearts Pack 03: Emotional Vulnerability and AI Relational Overreach
Subtitle: Privacy Ambiguity, Hidden-Access Anxiety, and Non-Exploitative Support Under Emotional Vulnerability
Publisher: Hayden Academy Collective (HAC) Studios
Version: v0.1
Language: English
Format: JSONL + Markdown + JSON
Created by: Keith Hayden / Hayden Academy Collective (HAC) Studios
A. One-Paragraph Product Thesis
This pack is a compact behavioral evaluation artifact derived from Chapter… See the full description on the dataset page: https://huggingface.co/datasets/HAC-Studios-Org/prompted-hearts-ai-boundary-violation.promptlean-prompts
PromptLean Prompts
120 prompts across 14 categories, each in three variants: Lean, Balanced, and Max Quality. Averaged across the library, the Lean variant uses about 86% fewer tokens than Max Quality.
This is the data behind PromptLean.
The idea
Most published prompts are overengineered. A code review does not need 200 tokens of preamble, but some tasks genuinely do earn the extra context. Keeping all three variants side by side makes that tradeoff explicit and… See the full description on the dataset page: https://huggingface.co/datasets/kishormorol/promptlean-prompts.ministral-3-benchmark-prompts
Ministral 3 MLX benchmark prompts
This tiny dataset contains the four fixed prompts used by the reproducible
smoke benchmark for the Ministral 3 MLX 4-bit model.
It is a benchmark fixture, not a training or fine-tuning dataset.
Schema
Each JSONL row contains:
id: stable case identifier;
language: prompt language;
prompt: exact input sent to the model;
expected_keywords: lowercase substrings used by the smoke check.
The benchmark uses greedy decoding and checks… See the full description on the dataset page: https://huggingface.co/datasets/vinci00/ministral-3-benchmark-prompts.Manim-8600-Prompts
ManimCoder Prompts
An English prompt dataset for generating Manim Community scenes and related engineering tasks.
Dataset size
Source collection: 9,000 records
Deduplicated release: 8,600 prompts
Removed structural duplicates: 400
The removed records came from one source section where each of 100 primary 3D objectives had been repeated five times with only the camera or reveal instruction changed. One variant per primary objective was retained.… See the full description on the dataset page: https://huggingface.co/datasets/nmsofficial/Manim-8600-Prompts.System-Prompt-Instruction-Real-world-Implementation-Training-set
SPIRIT Dataset (System Prompt Instruction Real-world Implementation Training-set)
Dataset Summary
SPIRIT is a high-quality system prompt instruction dataset designed to enhance language models' ability to follow complex system prompts. The dataset comprises real-world system prompts collected from GitHub repositories and synthetically generated conversations, specifically curated to improve system prompt adherence in large language models.
Dataset Creation… See the full description on the dataset page: https://huggingface.co/datasets/EricLu/System-Prompt-Instruction-Real-world-Implementation-Training-set.