datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
PhysicsEval
PhysicsEval Dataset
To enable large-scale evaluation and training of reasoning-capable language models in physics, we curated a comprehensive dataset of 19,609 annotated problems, sourced from 20 authoritative physics textbooks and verified educational websites.
Construction
The dataset spans 19 different categories, including Mechanics, Thermodynamics, Electromagnetism, Waves, Optics, Relativity, and Quantum Physics.
Each problem is processed through the following… See the full description on the dataset page: https://huggingface.co/datasets/IUTVanguard/PhysicsEval.Physics-RW
Physics-RW for lmms-eval
Normalized 1,135-example English Physics-RW classification benchmark and official videos.
Source: https://www.modelscope.cn/datasets/pengyz/Physics-RW. The source dataset license is preserved.
OpenSciReasoning-Physics-20K
OpenSciReasoning-Physics-20K
Three-domain release derived from nvidia/OpenScienceReasoning-2 for
domain-specific reasoner training and cross-domain transfer experiments.
Each row preserves the stable source_row_id and has exactly one mutually
exclusive domain value: PHYSICS. Domain acceptance was checked from the
question and choices with two independent question-only verifiers; answer and
source-ID gates were also replayed.
The audit records list any remaining source-output… See the full description on the dataset page: https://huggingface.co/datasets/TerryJCZhang/OpenSciReasoning-Physics-20K.physics-reasoning-dataset
📚 Flux Physics Reasoning Dataset
This dataset contains detailed physics reasoning scenarios designed to train Small Language Models (SLMs) and Liquid Neural Networks in physical intuition.
📄 Format
The dataset is provided in Parquet format (train.parquet) for efficient loading. Each row contains:
prompt: The physics question or scenario description.
answer: The correct physical explanation and answer.
concept: The underlying physics principle (e.g., "Conservation of… See the full description on the dataset page: https://huggingface.co/datasets/convaiinnovations/physics-reasoning-dataset.physcorp-a
PhysCorp-A — Audited Physics Training Corpus (6,432 records)
The audit-clean training corpus released alongside the Physics-R1 paper (NeurIPS 2026 D&B Track submission). Subset of the 14,294-record pre-audit pool that survives the joint two-stage contamination audit against all six paper-canonical eval splits.
Lineage
14,294 pre-audit pool (PhysCorp-pre-audit) aggregating nine source families.
Construction audit (Stage-1 5-gram Jaccard ≥ 0.4 union Stage-2… See the full description on the dataset page: https://huggingface.co/datasets/physics-r1-anonymous/physcorp-a.physics-verified
PHYSICS-Verified
PHYSICS-Verified is a benchmark of 1,109 PhD-qualifying-exam physics problems with 2,803 scored answers, covering six core areas of physics. Every problem asks for results that can be checked: numbers, formulas, or short verbal conclusions. Each answer has been checked against its reference solution.
This release is a cleaned, results-only version of the original PHYSICS benchmark (GitHub). Problems that required a proof, explanation, or drawing were removed, as… See the full description on the dataset page: https://huggingface.co/datasets/yale-nlp/physics-verified.Thai-Physics-Data-40KThai-Physics-Data is a Thai-Based physics data with more than 40k lines of data.
Data Sources:
ArtifactAI/arxiv-physics-instruct-tune-30k (CC BY-NC 2.0)
camel-ai/physics
How to load Data (Hugging Face)
from datasets import load_dataset
Thai_Physics_Data = load_dataset("Kongongong/Thai-Physics-Data-40K")
Thai_Physics_Data = Thai_Physics_Data['train']
def format_data():
....
data =[]
format_data()
data = Dataset.from_dict({"text": data})
How to load Data (CSV)
from… See the full description on the dataset page: https://huggingface.co/datasets/Kongongong/Thai-Physics-Data-40K.Physics-DPO-Dataset
Dataset Summary
Physics-DPO is a meticulously crafted, synthetic dataset of preference pairs designed to enhance the physics reasoning and problem-solving capabilities of Large Language Models (LLMs). Generated using a sophisticated, multi-stage pipeline, this dataset provides high-quality data for Direct Preference Optimization (DPO) and similar preference alignment techniques.
Each instance consists of a challenging physics problem (prompt), a high-quality, expert-level solution… See the full description on the dataset page: https://huggingface.co/datasets/yusufbaykaloglu/Physics-DPO-Dataset.Dense-Information-Science-Physics-Dataset
Dense Information With Multiple Fine-tuned Variations
This dataaset has multiple for each input to learn how to express the same answer in different ways
Dataset Structure
The dataset contains two columns:
Column
Description
input
A science or quantum-physics question
output
A conversational answer to the question
Example:
{
"input": "What is quantum entanglement?",
"output": "Quantum entanglement is when two quantum systems share one… See the full description on the dataset page: https://huggingface.co/datasets/StarpowerTechnology/Dense-Information-Science-Physics-Dataset.NCERT_Physics_11thhigh-school-physicsquantum-hardware-device-physics
Neura Parse — Quantum Hardware Device Physics: Qubit Design, Coherence, Control & Scaling
A physics- and engineering-deep vertical on how qubits are built, controlled, and scaled across superconducting, trapped-ion, neutral-atom, and spin modalities (plus emerging erasure/biased-noise qubits). Device-physics derivations, coherence-limit analyses, control-stack engineering, and 2025-2026 scaling/interconnect work, with QuTiP/scqubits simulation context — expanding the general… See the full description on the dataset page: https://huggingface.co/datasets/Neura-parse/quantum-hardware-device-physics.physics-qa
Physics Q&A — Multi-Level Explanations
502 question-answer pairs generated from recent physics papers (arXiv 2024–2026),
covering 5 subfields across 83 papers.
Each paper is explained at 6 audience levels, each as a focused Q/A pair:
Level
Audience
physicist
Expert with equations and notation
undergrad_science
Undergraduate with key equations
high_schooler
High school student, intuitive
humanities_student
No math, analogies only
five_year_old
Child-friendly… See the full description on the dataset page: https://huggingface.co/datasets/planetoid-reader/physics-qa.physr1corp
PhysR1Corp — Closed-form RL Training Pool (2,268 records)
The closed-form (numeric / MCQ-gradable) RL training pool used by Physics-R1 (NeurIPS 2026 D&B Track submission). Carved out of PhysCorp-A (the audited 6,432-record corpus) by dropping open-ended questions, then decontaminated against MMMU-Pro Physics (−87 records) and against PhyX-mini + PhysUniBench-en at cos ≥ 0.85 (−78 additional records: 69 PhyX-mini near-duplicates and 9 PhysUniBench-en template duplicates). One… See the full description on the dataset page: https://huggingface.co/datasets/physics-r1-anonymous/physr1corp.NCERT_Physics_12thsyth-lasers-physics-semiconductors
syth-lasers-physics-semiconductors
Synthetic Q&A dataset on physics of lasers, diodes and semi conductors., generated with SDGS (Synthetic Dataset Generation Suite).
Dataset Details
Metric
Value
Topic
physics of lasers, diodes and semi conductors.
Total Q&A Pairs
105
Valid Pairs
105
Provider/Model
ollama/default
Generation Cost
Metric
Value
Prompt Tokens
52,657
Completion Tokens
56,455
Total Tokens
109,112
GPU Energy… See the full description on the dataset page: https://huggingface.co/datasets/Kylan12/syth-lasers-physics-semiconductors.Math-physics-dataset-sq
Physics and Math Problems Dataset
This repository contains a dataset of 2,600 physics and math problems in Albanian. The dataset is designed to support various NLP tasks and educational applications.
Dataset Overview
Problems: 2,666
Language: Albanian
Fields:
algebra lineare: 178
analiza matematike: 176
gjeometri diferenciale: 178
topologji: 179
teoria e numrave: 177
ekuacionet diferenciale: 178
fizika klasike: 177
mekanika kuantike: 179
elektromagnetizmi: 178… See the full description on the dataset page: https://huggingface.co/datasets/LTS-VVE/Math-physics-dataset-sq.physics_arabic_CoT_10K
Dataset Summary
The Arabic Physics Problem Dataset is a curated collection of Arabic-language physics exam questions.Each record contains a question written in Modern Standard Arabic, the correct answer, a detailed explanation, and the related physics concept.This dataset is part of the Mobiusi multilingual STEM education initiative, created to support scientific reasoning, question answering, and educational AI applications in Arabic-speaking contexts.
Each sample follows a… See the full description on the dataset page: https://huggingface.co/datasets/Mobiusi/physics_arabic_CoT_10K.brazilian-math-physics-qa
Brazilian Math & Physics QA
English | Português do Brasil
English
Summary
Brazilian Portuguese question-answer pairs covering mathematics, physics, chemistry, and related educational subjects. Each record contains a user question and an assistant answer in chat/SFT format.
Examples: 19.082
Train: 18.148
Validation: 934
Language: Brazilian Portuguese (pt-BR)
Schema
{"id":"qa_...","subject":"fisica","category":"mecanica-geral"… See the full description on the dataset page: https://huggingface.co/datasets/artificialguybr/brazilian-math-physics-qa.brazilian-math-physics-qa-vision
Brazilian Math & Physics QA — Image Dependent
English | Português do Brasil
English
Summary
Brazilian Portuguese educational question-answer pairs whose problem statement or solution depends on one or more images.
Examples: 3,808
Referenced image URLs: 5,094 unique
Language: Brazilian Portuguese (pt-BR)
Schema
{"id":"vqa_...","subject":"matematica","category":"geometria","title":"...","messages":[{"role":"user","content":"...… See the full description on the dataset page: https://huggingface.co/datasets/artificialguybr/brazilian-math-physics-qa-vision.academia-physics-office-hours
This Dialogue
Comprised of fictitious examples of dialogues between a physics professor and a student during office hours. Check out the example below:
"id":1,
"description":"Understanding the concept of velocity",
"dialogue":"Student: Professor, I'm having trouble understanding the concept of velocity. Could you please explain it to me?\n\nProfessor: Of course! Velocity is a fundamental concept in physics that describes the rate of change of an object's position with respect to… See the full description on the dataset page: https://huggingface.co/datasets/FunDialogues/academia-physics-office-hours.pocket-physicsNCERT_Physics_12thphysics-30k-demo
Computational & Quantitative Sciences Q&A — Multi-Level Explanations
24 question-answer pairs generated from recent papers (arXiv 2024–2026),
covering 6 subfields across 6 papers.
Each paper is explained at 4 depth levels, each as a focused Q/A pair:
Level
Description
L1
Intuitive / Phenomenological — what is happening, plain language, analogies, no equations
L2
Conceptual / Structural — key components, pipeline/steps, minimal formalism
L3
Mechanistic / Formal —… See the full description on the dataset page: https://huggingface.co/datasets/planetoid-reader/physics-30k-demo.physcorp-pre-audit
PhysCorp Pre-Audit Raw Pool (14,294 records)
The pre-audit master corpus released alongside the Physics-R1 paper (NeurIPS 2026 D&B Track submission), aggregating nine source families before contamination audit. Released so users can re-run the audit at different thresholds or against new external benchmarks.
Source breakdown
Source
Records
License
Tier
UGPhysics
5,520
CC BY-NC-SA 4.0
Tier-2
OpenStax College + University Physics
2,381
CC BY 4.0
Tier-1… See the full description on the dataset page: https://huggingface.co/datasets/physics-r1-anonymous/physcorp-pre-audit.cleand_moremilk_CoT_Reasoning_Quantom_Physics_And_Computing元データ: https://huggingface.co/datasets/moremilk/CoT_Reasoning_Quantom_Physics_And_Computing
使用したコード: https://github.com/LLMTeamAkiyama/0-data_prepare/tree/master/src/CoT_Reasoning_Quantom_Physics_And_Computing
データ件数: 2,862
平均トークン数: 1,110
最大トークン数: 2,334
合計トークン数: 3,175,666
ファイル形式: JSONL
ファイル分割数: 1
合計ファイルサイズ: 15.5 MB
加工内容：
メタデータ列の解析と新列生成: metadata列（辞書型）を解析し、その中のreasoningをthought列に、difficultyをdifficulty列に展開しました。解析に失敗した行は除外されました。また、元のmetadata列は削除されました。
難易度によるフィルタリング:… See the full description on the dataset page: https://huggingface.co/datasets/LLMTeamAkiyama/cleand_moremilk_CoT_Reasoning_Quantom_Physics_And_Computing.beir_cqadupstack_physics_test
beir_cqadupstack_physics_test
BEIR CQADupStack/physics test split
Field
Value
Benchmark
beir
Sub-benchmark
cqadupstack_physics
Type
retrieval
Items
1039
Exported from Langfuse.
quantum_physicsphysics-5-4-mini-high
rl-rag-2/physics-5-4-mini-high
99 / 100 constructed queries (1 skipped due to parse errors).
Fields
seed_thread_id
context
setup
core_question
sub_questions
rubrics
seed_query
seed_links
claims
claims_usage: token usage to construct claims with OpenAI built-in web search
question_usage: token usage to construct research context, setup, core question, sub questions, rubrics from claims. No search.
total_cost: total cost USD of claims + question construction
physics_squad_trial
Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
Dataset Details
Dataset Description
Curated by: [More Information Needed]
Funded by [optional]: [More Information Needed]
Shared by [optional]: [More Information Needed]
Language(s) (NLP): [More Information Needed]
License: [More Information Needed]
Dataset Sources [optional]
Repository: [More… See the full description on the dataset page: https://huggingface.co/datasets/vinaypri/physics_squad_trial.