Datasets:

jablonkagroup
/

chempile-paper

fn string	text string	doi string	title string	authors string	__index_level_0__ int64
10.48550_arXiv.0708.1447	###### Abstract The non-specific adhesion of spherical particles to a cell substrate is analyzed in a parallel plate flow chamber, addressing the effect of the particle size. Differently from other experiments, the total volume of the injected particles has been fixed, rather than the total number of particles, as the...	10.48550/arXiv.0708.1447	On the adhesion of particles to a cell layer under flow	F. Gentile, A. Granaldi, P. Decuzzi	5,795
10.48550_arXiv.0807.5119	###### Abstract We have carried out density functional theory based calculations of the size dependent formation energy and geometry of Pt islands on Ru, to model Pt-Ru nanocatalysts which have been recently proposed as fuel cell anode. The Pt islands are found to prefer two-dimensional structures. Furthermore, a mono...	10.48550/arXiv.0807.5119	Formation of Pt islets on facets of Ru nanoparticles: a first-principles study	Marisol Alcántara Ortigoza, Sergey Stolbov, Talat Rahman	4,712
10.48550_arXiv.0711.4041	###### Abstract We study electromagnetic properties of periodic composite structures, such as photonic crystals, involving lossy components. We show that in many cases a properly designed periodic structure can dramatically suppress the losses associated with the absorptive component, while preserving or even enhancin...	10.48550/arXiv.0711.4041	Absorption suppression in photonic crystals	A. Figotin, I. Vitebskiy	2,543
10.48550_arXiv.1605.06548	###### Abstract The adiabatic elastic modulus is often useful in the high frequency response of materials. Unfortunately, it can be much more difficult to directly measure the adiabatic elastic modulus of material than the isothermal elastic modulus. We derive the relationship between the adiabatic and isothermal elas...	10.48550/arXiv.1605.06548	Isothermal and Adiabatic Elastic Tensors	Michael J. Waters, Andrew W. Bielawski	4,751
10.48550_arXiv.1610.09804	###### Abstract Dirac-like electronic states are the main engines powering the tremendous advances in research of graphene, topological insulators and other materials with these states. Zero effective mass, high carrier mobility and numerous applications are some consequences of linear dispersion that distinguishes Di...	10.48550/arXiv.1610.09804	Pyramids and cootie catchers: new massless fermions in 2D materials	Vladimir Damljanovic, Rados Gajic, Igor Popov	2,087
10.48550_arXiv.1101.4248	###### Abstract We present an _ab initio_ theory of core- and valence resonant inelastic x-ray scattering (RIXS) based on a real-space multiple scattering Green's function formalism and a quasi-boson model Hamiltonian. Simplifying assumptions are made which lead to an approximation of the RIXS spectrum in terms of a c...	10.48550/arXiv.1101.4248	Real Space Green's Function Approach to RIXS	J. J. Kas, J. J. Rehr, J. A. Soininen, P. Glatzel	2,328
10.48550_arXiv.1512.07791	###### Abstract In this work, the elastic and thermodynamic properties of Pt${}_{3}$Al under high pressure are investigated using density functional theory within the generalized gradient approximation. The results of bulk modulus and elastic constants at zero pressure are in good agreement with the available theore...	10.48550/arXiv.1512.07791	Theoretical study of the elastic and thermodynamic properties of Pt$_{3}$Al with the L1$_{2}$ structure under high pressure	N. Wei, Ch. Zhang, S. Hou	5,663
10.48550_arXiv.1005.1115	###### Abstract The electronic structure and magnetic coupling properties of rare-earth metals (Gd, Nd) doped ZnO have been investigated using first-principles methods. We show that the magnetic coupling between Gd or Nd ions in the nearest neighbor sites is ferromagnetic. The stability of the ferromagnetic coupling b...	10.48550/arXiv.1005.1115	Magnetic coupling properties of rare-earth metals (Gd, Nd) doped ZnO: first-principles calculations	Hongliang Shi, Ping Zhang, Shu-Shen Li, Jian-Bai Xia	3,426
10.48550_arXiv.1404.1658	## Edge step normalized Co-K edge XANES spectra of ceramic CTO sample. Co metal is used for photon energy calibration. Cobalt oxide (CoO and CoF${}_{3}$) standards are shown together to obtain Co valency in CTO. ## LCF fit of CTO raw data with CoO and CoF${}_{3}$ standard samples and linear dependence of oxidati...	10.48550/arXiv.1404.1658	Coexistence of Co3+ and Co2+ in ceramic Co3TeO6; XANES, Magnetization and first principle study	Harishchandra Singh, Haranath Ghosh, T. V. Chandrasekhar Rao, A. K. Sinha	4,831
10.48550_arXiv.1005.3344	"###### Abstract\n\nWe have investigated plasma-surface interactions with molecular dynamics (MD) si(...TRUNCATED)	10.48550/arXiv.1005.3344	Extension of the simulation code ACAT to treat real atomic positions	Arimichi Takayama, Seiki Saito, Atsushi M. Ito, Takahiro Kenmotsu, Hiroaki Nakamura	2,127

End of preview. Expand in Data Studio

ChemPile-Paper

A comprehensive collection of scientific literature spanning academic papers and preprints focused on chemistry and related fields

📋 Dataset Summary

ChemPile-Paper serves as a resource for cutting-edge applications of chemical knowledge and reasoning, containing curated papers from diverse repositories. This dataset represents a comprehensive collection of scientific literature spanning academic papers and preprints, all focused on chemistry and related fields.

📊 Dataset Statistics

Subset	Tokens	Documents	Description
ArXiv Cond-Mat Materials Science	35.9M	5,94K	Materials science papers
ArXiv Physics Chemical Physics	62.3M	6,75K	Chemical physics papers
bioRxiv	82.6M	60,3K	Biology preprints
medRxiv	451M	15,8K	Health sciences preprints
ChemRxiv	210M	28,9K	Chemistry community preprints
EuroPMC Chemistry Abstracts	3.3B	10.4M	Scientific literature abstracts
EuroPMC Chemistry Papers	10B	1.2M	Scientific literature full articles
Total	~13.9B	~11.7M	Chemical scientific literature

🗂️ Dataset Configurations

The dataset includes different subsets available as Hugging Face configurations:

arxiv-cond-mat.mtrl-sci_processed-default
arxiv-physics.chem-ph_processed-default
biorxiv_processed-default
chemrxiv_processed-default
euro_pmc_chemistry_abstracts-default
euro_pmc_chemistry_papers-default
medrxiv_processed-default

📜 License

All content is released under the CC BY-NC-ND 4.0 license, which allows for:

✅ Non-commercial use
✅ Sharing and redistribution
⚠️ Attribution required
❌ No derivatives allowed

📖 Dataset Details

📚 ArXiv Subsets

Sources:

Condensed Matter > Materials Science (arxiv-cond-mat.mtrl-sci_processed-default)
Physics > Chemical Physics (arxiv-physics.chem-ph_processed-default)

Coverage: Academic papers from ArXiv in materials science and chemical physics

Extraction Method: Articles filtered by field using PaperScraper package for PDF download and processing

Fields:

fn: ArXiv identifier (e.g., 10.48550_arXiv.0708.1447)
text: Parsed text of the article
doi: DOI of the article (if available)
title: Article title
authors: Article authors
index: Document identifier

Statistics:

Materials Science: 35.9M tokens across 5,940 documents
Chemical Physics: 62.3M tokens across 6,750 documents

🧬 bioRxiv and medRxiv

Sources:

bioRxiv - Biology preprint repository
medRxiv - Health sciences preprint repository

Coverage: Preprints in biology and health sciences with chemistry relevance

Extraction Method: PaperScraper package for DOI-based retrieval, processed with Nougat for text extraction

Fields:

fn: Unique identifier (e.g., 014597_file10)
text: Full text content extracted via Nougat

Statistics:

bioRxiv: 82.6M tokens across 60,300 documents
medRxiv: 451M tokens across 15,800 documents

⚗️ ChemRxiv

Source: ChemRxiv - Preprint server for the global chemistry community

Coverage: Chemistry preprints from the community

Extraction Method: PaperScraper for DOI-based retrieval, processed with Nougat for text extraction

Fields:

fn: Unique identifier (e.g., 10.26434_chemrxiv-2022-cgnf5)
text: Full text content extracted via Nougat
doi: DOI of the article (if available)
title: Article title
authors: Article authors
license: Preprint license (e.g., CC BY-NC 4.0)
published_url: Publication URL
index: Document identifier

Statistics: 210M tokens across 28,900 documents

🔬 EuroPMC Filtered Papers

Source: EuroPMC - 27 million abstracts and 5 million full-text articles

Coverage: Chemistry-related scientific papers filtered from comprehensive medical literature

Extraction Method:

BERT-based multilabel classifier trained on CAMEL datasets (20,000 examples per discipline)
Validated against FineWebMath annotations (F1-score ~0.77 on 150 manually annotated entries)
Analysis of first five 512-token chunks per document with 50-token overlaps

Quality Control:

Postprocessing to remove non-chemical content (authors, acknowledgments, page numbers)
Chemistry-specific content identification and filtering

Fields:

pmcid: PubMed Central identifier
pmid: PubMed identifier
topic: Main classification topic (e.g., "Chemistry", "Physics", "Biology")
confidence: Classification confidence score
class_distribution: Multilabel classification distribution
text: Full article text content

Statistics:

Abstracts: 3.3B tokens across 10.4M documents
Full Papers: 10B tokens across 1.2M documents

🚀 Quick Start

from datasets import load_dataset, get_dataset_config_names

# List all available configurations
configs = get_dataset_config_names("jablonkagroup/chempile-paper")
print(f"Available configs: {configs}")
# ['arxiv-cond-mat.mtrl-sci_processed-default', 'arxiv-physics.chem-ph_processed-default', 
#  'biorxiv_processed-default', 'chemrxiv_processed-default', 'euro_pmc_chemistry_abstracts-default',
#  'euro_pmc_chemistry_papers-default', 'medrxiv_processed-default']

# Load a specific subset
dataset = load_dataset("jablonkagroup/chempile-paper", name="arxiv-cond-mat.mtrl-sci_processed-default")

print(dataset)
# DatasetDict({
#     train: Dataset({
#         features: ['fn', 'text', 'doi', 'title', 'authors', '__index_level_0__'],
#         num_rows: 5899
#     })
#     test: Dataset({
#         features: ['fn', 'text', 'doi', 'title', 'authors', '__index_level_0__'],
#         num_rows: 328
#     })
#     val: Dataset({
#         features: ['fn', 'text', 'doi', 'title', 'authors', '__index_level_0__'],
#         num_rows: 328
#     })
# })

# Access a sample
sample = dataset['train'][0]
print(f"Sample ID: {sample['fn']}")
print(f"Sample text: {sample['text'][:200]}...")

🎯 Use Cases

🤖 Language Model Training: Pre-training or fine-tuning models for chemistry domain with cutting-edge research
🔬 Research Intelligence: Building systems for scientific literature analysis and discovery
🔍 Information Retrieval: Advanced chemistry knowledge base construction from research literature
📝 Content Generation: Automated scientific writing and research synthesis
🧠 Domain Adaptation: Adapting models to cutting-edge chemical research and terminology

⚠️ Limitations & Considerations

Language: Primarily English (monolingual dataset)
Scope: Focused on published research; may include technical jargon and advanced concepts
Quality: Variable quality across sources; some OCR errors possible in older papers
Bias: Reflects biases present in scientific publishing and academic literature
License: No derivatives allowed due to CC BY-NC-ND 4.0 license
Recency: Content reflects publication dates; cutting-edge developments may not be included

🛠️ Data Processing Pipeline

Collection: Automated scraping from academic repositories and databases
Filtering: BERT-based classification for chemistry relevance
Extraction: PDF processing with PaperScraper and Nougat OCR
Quality Control: Automated filtering and expert validation
Standardization: Consistent formatting and metadata extraction
Validation: Train/validation/test splits and quality checks

🏗️ ChemPile Collection

This dataset is part of the ChemPile collection, a comprehensive open dataset containing over 75 billion tokens of curated chemical data for training and evaluating general-purpose models in the chemical sciences.

Collection Overview

📊 Scale: 75+ billion tokens across multiple modalities
🧬 Modalities: Structured representations (SMILES, SELFIES, IUPAC, InChI), scientific text, executable code, and molecular images
🎯 Design: Integrates foundational educational knowledge with specialized scientific literature
🔬 Curation: Extensive expert curation and validation
📈 Benchmarking: Standardized train/validation/test splits for robust evaluation
🌐 Availability: Openly released via Hugging Face

📄 Citation

If you use this dataset in your research, please cite:

@article{mirza2025chempile0,
  title   = {ChemPile: A 250GB Diverse and Curated Dataset for Chemical Foundation Models},
  author  = {Adrian Mirza and Nawaf Alampara and Martiño Ríos-García and others},
  year    = {2025},
  journal = {arXiv preprint arXiv:2505.12534}
}