datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
4B-think-reason-only.stride-train1-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
4B-think-reason-only.stride-train1-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
A pre-tokenized, tag-wrapped variant of JackHsieh/4B-think-reason-only.stride-train1-test32.k-8.statml-arxiv.qwen3-ids.
The source thoughts are two parts: a <think> block, then a few dense sentences that the
generator wrapped in a literal <thought>…</thought> block (plain text, not special tokens).
Only the part after </think> becomes the VALUE, with every <thought> / </thought> tag
removed… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/4B-think-reason-only.stride-train1-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained.4B-think-paragraph.stride-train1-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
4B-think-paragraph.stride-train1-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
A pre-tokenized, tag-wrapped variant of JackHsieh/4B-think-paragraph.stride-train1-test32.k-8.statml-arxiv.qwen3-ids.
The source thoughts are two parts: a <think> block, then a single paragraph of dense
reasoning about the immediate continuation. Only the part after </think> — the paragraph —
becomes the VALUE. The reasoning inside the think block is dropped.
The VALUE is capped at 512 tokens.… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/4B-think-paragraph.stride-train1-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained.CommonsenseQA-Explained-by-ChatGPTThis is a dataset with explanations from ChatGPT for the correct and incorrect answers in CommonsenseQA. The explanations are generated by prompting ChatGPT with answer keys and in-context examples. We expect this dataset to be an useful source for understanding the commonsense reasoning ability of LLMs or training other LMs.
4B-Instruct-reason-only.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
4B-Instruct-reason-only.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
Tokenized, tag-wrapped form of JackHsieh/4B-reason-only.rule-r-1.0-k-8.L-512.statml-arxiv.
Each thought is wrapped as
<|note|>
This is a hint about a span that appears later in this document. KEY is the text immediately before that span; VALUE is a note about what might come next.
KEY: <last 8 prefix tokens>
VALUE: <thought>
<|/note|>
and stored both as text (thought_text) and as
Qwen/Qwen3-4B-Base… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/4B-Instruct-reason-only.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained.EDD_explain_dataxpo-explainer-dataexplainxkcd
Dataset Card for Explain XKCD
Dataset Summary
An import of the Explain XKCD wiki (up to March 30, 2024), containing explanations for 2668 xkcd comics.
Field
Type
Description
number
string
The xkcd comic’s sequential number (e.g. "3", "1000").
date
string
Publication date of the comic, in “Month Day, Year” format (e.g. "September 30, 2005").
title
string
The displayed title of the comic (e.g. "Island (sketch)").
before
string
Any original or archived… See the full description on the dataset page: https://huggingface.co/datasets/dchichkov/explainxkcd.eli5_rlhf_explainlikeim5
ELI5 paired
This is a processed version of the eli5 dataset.
Compared to "eli5_rlhf", this dataset contains only QA pairs from the train split of the eli5 dataset and only from the subreddit explainlikeimfive.
Furthermore, the function
def get_question(example):
title = example["title"]
selftext = example["selftext"]
if selftext:
if selftext[-1] not in [".", "?", "!"]:
seperator = ". "
else:
seperator = " "
question = title… See the full description on the dataset page: https://huggingface.co/datasets/vincentmin/eli5_rlhf_explainlikeim5.multilingual_toxicity_explained
Multilingual and Explainable Toxicity
We explained the toxic part of our multilingual ParaDetox dataset utilizing GPT-4 (May, 2024) with the following prompt:
Please analyze the provided sentence using the structure below to identify elements of
toxicity and suggest improvements, when I tell you, use words from the keywords list (can be
more than one word!):
keywords = [Neutral, Informative, Casual, Assertive, Dismissive, Condescending,
Friendly, Commanding, Instructive… See the full description on the dataset page: https://huggingface.co/datasets/textdetox/multilingual_toxicity_explained.ARC-Challenge-Explained-by-ChatGPTThis is a dataset with explanations from ChatGPT for the correct and incorrect answers in ARC Challenge. The explanations are generated by prompting ChatGPT with answer keys and in-context examples. We expect this dataset to be an useful source for understanding the commonsense reasoning ability of LLMs or training other LMs.
cli-commands-explained
Overview
This dataset is a collection of 16,098 command line instructions sourced from Commandlinefu and Cheatsheets. It includes an array of commands, each with an id, title, description, date, url to source, author, votes, and flag indicating if the description is AI generated. The descriptions are primarily authored by the original contributors, for entries where descriptions were absent, they have been generated using NeuralBeagle14-7B. Out of the total entries, 10,039… See the full description on the dataset page: https://huggingface.co/datasets/b-mc2/cli-commands-explained.4B-Instruct-distill-jx739p0v.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
4B-Instruct-distill-jx739p0v.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
A pre-tokenized, tag-wrapped variant of JackHsieh/4B-Instruct-distill-jx739p0v.stride-1.k-8.statml-arxiv.qwen3-ids. Each thought_text is wrapped as
<|note|>
This is a hint about a span that appears later in this document. KEY is the text immediately before that span; VALUE is a note about what might come next.
KEY: {last 8 prefix tokens}
VALUE: {thought_text}
<|/note|>
and stored both as text… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/4B-Instruct-distill-jx739p0v.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained.labeled-multiple-choice-explained-mistral-reasoningARC-Easy-Explained-by-ChatGPTThis is a dataset with explanations from ChatGPT for the correct and incorrect answers in ARC-Easy. The explanations are generated by prompting ChatGPT with answer keys and in-context examples. We expect this dataset to be an useful source for understanding the commonsense reasoning ability of LLMs or training other LMs.
labeled-multiple-choice-explained-falcon-reasoning4B-ranked-v7.rule-stride-train4-test32.k-8.L-4096.statml-arxiv.qwen3-ids.kv-tags-explainedchatgpt-explain-sentiment
Explain Sentiment
Generated using ChatGPT3.5 on Malaysian tweets, notebooks at https://github.com/mesolitica/malaysian-dataset/tree/master/sentiment/chatgpt3.5-sentiment
sentiment.jsonl, 162902 rows, 86 MB
Example data
{'sentiment': 'negative',
'explain_en': 'The text is negative because it contains an angry tone and disrespectful language towards someone named Amzar.',
'explain_ms': 'Teks ini negatif kerana mengandungi nada marah dan bahasa yang tidak sopan… See the full description on the dataset page: https://huggingface.co/datasets/mesolitica/chatgpt-explain-sentiment.joke_explaination
Dataset Card for Dataset Name
Dataset Summary
Corpus for testing whether your LLM can explain the joke well. But this is a rather small dataset, if someone can point to a larger ones would be very nice.
Languages
English
Dataset Structure
Data Fields
url : link to the explaination
joke : the original joke
explaination : the explaination of the joke
Data Splits
Since its so small, there's no splits just like gsm8k
4B-Instruct-distill-3rkrz2vo.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
4B-Instruct-distill-3rkrz2vo.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
A pre-tokenized, tag-wrapped variant of JackHsieh/4B-Instruct-distill-3rkrz2vo.stride-1.k-8.statml-arxiv.qwen3-ids. Each thought_text is wrapped as
<|note|>
This is a hint about a span that appears later in this document. KEY is the text immediately before that span; VALUE is a note about what might come next.
KEY: {last 8 prefix tokens}
VALUE: {thought_text}
<|/note|>
and stored both as text… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/4B-Instruct-distill-3rkrz2vo.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained.spam.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
spam.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
A length-matched null control for JackHsieh/luna-reason-only.stride-train8-test32.k-8.statml-arxiv.qwen3-ids, stored directly in the
kv-tags-explained training format (there is no separate base repo). Every thought is this
one sentence, repeated:
We are thinking hard about what comes in the next 8 tokens by reasoning correctly and carefully about what comes before it in the document.
It is fluent, on-topic and… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/spam.stride-1.k-8.statml-arxiv.qwen3-ids.kv-tags-explained.repro-time-series-saliency-maps-explaining-models-across-multiple-domains-traces
Agent traces
Agent sessions published from a Trackio Logbook.
4B-ranked-v9b.stride-train2-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
4B-ranked-v9b.stride-train2-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
A pre-tokenized, tag-wrapped variant of JackHsieh/4B-ranked-v9b.stride-train2-test32.k-8.statml-arxiv.qwen3-ids.
The source thoughts are two parts: a <think> block, then a ranked list of candidate
continuations. Only the part after </think> — the list — becomes the VALUE. The reasoning
inside the think block is dropped.
The VALUE is capped at 512 tokens.
4.38% of rows have no usable list: the… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/4B-ranked-v9b.stride-train2-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained.autotrain-data-code-explainer
AutoTrain Dataset for project: code-explainer
Dataset Description
This dataset has been automatically processed by AutoTrain for project code-explainer.
Languages
The BCP-47 code for the dataset's language is unk.
Dataset Structure
Data Instances
A sample from this dataset looks as follows:
[
{
"text": "def upload_to_s3(local_file, bucket, s3_file):\n ## This function is responsible for uploading the file into the S3 bucket using… See the full description on the dataset page: https://huggingface.co/datasets/sagard21/autotrain-data-code-explainer.pawn-loans-explained
Pawn loans explained
King Gold & Pawn presents a Canonical Explainer on Pawn Loans Pawn Loans Explained: A Comprehensive Research Guide Introduction to Pawn Loans A pawn loan represents one of the oldest forms of secured lending, with a history tracing back thousands of years across various civilizations. At its core, a pawn loan is a non-recourse loan provided to a borrower in exchange for an item of value, known as collateral. The distinguishing characteristic of a pawn loan is… See the full description on the dataset page: https://huggingface.co/datasets/CollateralAnalytics/pawn-loans-explained.luna-reason-only.stride-train8-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
luna-reason-only.stride-train8-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained
A pre-tokenized, tag-wrapped variant of JackHsieh/luna-reason-only.stride-train8-test32.k-8.statml-arxiv.qwen3-ids. Each thought_text is wrapped as
<|note|>
This is a hint about a span that appears later in this document. KEY is the text immediately before that span; VALUE is a note about what might come next.
KEY: {last 8 prefix tokens}
VALUE: {thought_text}
<|/note|>
and stored both as text… See the full description on the dataset page: https://huggingface.co/datasets/JackHsieh/luna-reason-only.stride-train8-test32.k-8.statml-arxiv.qwen3-ids.kv-tags-explained.vihsd-explainable-dpo
vihsd-explainable-dpo
DPO preference dataset derived from vominhmanh/vihsd-explainable for Direct Preference Optimization (DPO).
Each example is a preference pair (chosen vs rejected) for the same prompt.
Schema (per example):
prompt (string): the original SFT prompt for Vietnamese moderation (user instruction).
chosen (string): JSON string with keys explanation, evidence, label — preferred (longer/more informative) explanation.
rejected (string): JSON string with keys explanation… See the full description on the dataset page: https://huggingface.co/datasets/vominhmanh/vihsd-explainable-dpo.labeled-multiple-choice-explainedThis dataset is based on under-tree/labeled-multiple-choice but using GPT-3.5-turbo to generate explanations for each answer option.
This was a very basic attempt to follow the Orca paper approach of a 'teacher' model to provide more context to some trivia questions.
Questions were deduplicated based on the question text.
I used the python library guidance to help generate the prompts. Below is the prompt template I used.
{{#role 'system'~}}
You are an AI assistant that helps people find… See the full description on the dataset page: https://huggingface.co/datasets/layoric/labeled-multiple-choice-explained.gvg-explainable-game-similarity
GVG Explainable Game Similarity Dataset 2026 (v0.1.0)
50 human-reviewed pairs of similar PC games (82 games) from Game V Game. Each row says why the two games are similar, what differs, and who each suits, in English and Chinese, and cites the two official Steam store records the claims were checked against (with capture dates).
Most similarity data says two games are alike and stops. This one is small on purpose: every row was read by a person against the current Steam records… See the full description on the dataset page: https://huggingface.co/datasets/lette2/gvg-explainable-game-similarity.sage-code-explain-ko-v1chess-explained-sf6-k12-curriculum-500k