Find research datasets worth reusing
Search datasets from major research repositories and use ShareScore to quickly assess how well each record supports discovery, access, and reuse.
76
datasets available to search
ShareScore release 0.9.0
Dataset results
76 results for “llm”
LLM-Based Knowledge Graph Construction from Materials Research Scientific Literature
<p>This dataset was constructed by creating a benchmark of 349 manually annotated triples, which were extracted from four different research articles in the field of materials science.</p>
Dataset: Towards a Knowledge Management Framework for LLM-Generated Personas in Collaborative Systems
Open the record for dataset details and reuse information.
Bug Reports - DBRD using LLM's
<p>This dataset contains the bug reports from JIRA, MongoDB, and Hyperledger converted into JSONL Format.</p>
Results and log of LLM-KG-Bench runs described in article "Benchmarking the Abilities of Large Language Models for RDF Knowledge Graph Creation and Comprehension: How Well Do LLMs Speak Turtle?", Frey et al. 2023
<p>Results and log of LLM-KG-Bench runs described in article ""Benchmarking the Abilities of Large Language Models for RDF Knowledge Graph Creation and Comprehension: How Well Do LLMs Speak Turtle?", Frey et al. 2023, to appear in proceedings for workshop DL4KG@ISWC 2023.</p> <p>For data on task FactExtractStatic please contact authors.</p>
OpenEvidence Safety and Comparative Efficacy of Four LLM's in Clinical Practice
ClinicalTrials.gov study NCT07199231. IPD Sharing: UNDECIDED. Countries: 1. Publications: 6.
Online repository for Paper "AgentFL: Scaling LLM-based Fault Localization to Project-Level Context"
<h3>Summary</h3> <p>This is the online repository for the arXiv paper "AgentFL: Scaling LLM-based Fault Localization to Project-Level Context".</p> <p>We also provide the results for the TSE'25 paper "SOAPFL: A Standard Operating Procedure for LLM-based Method-Level Fault Localization".</p> <h3>Environment</h3> <ul> <li><a href="https://github.com/rjust/defects4j/tree/v1.4.0">Defects4J-V1.4.0</a> (Note that the buggy items in V1.4.0 is identical with V1.2.0, we use V1.4.0 to avoid some problems in V1.2.0)</li> <li><a href="https://github.com/rjust/defects4j/tree/v2.0.0">Defects4J-V2.0.0</a></li> <li>Python version >= 3.8.5</li> </ul> <h3>Defects4J Mod</h3> <p>Before running AgentFL, please apply the files under the <code>AgentFL/Defects4J_mod</code> directory to modify your Defects4J V1.4.0/V2.0.0.</p> <h3>Run AgentFL</h3> <p>Set your own OpenAI API key in <code>AgentFL/camel/model_backend.py</code></p> <p>It's easy to run AgentFL for localizing a bug with the following command:</p> <p><code>python3 run.py --config <CONFIG_DIR> --version <D4J_VERSION> --project <PROJECT> --bugID <BUG_ID> --model <GPT_MODEL_NAME></code></p> <p>For example:</p> <p><code>python3 run.py --config Default --version 1.4.0 --project Closure --bugID 26 --model GPT_3_5_TURBO</code></p> <p>More configs can be seen under the directory <code>AgentFL/Config</code></p> <h3>Results</h3> <p>We release all of the results of AgentFL in the <code>AgentFL/Results</code> directory, including the evaluation results on Defects4J V1.4.0/V2.0.0 and the ablation study result.</p> <p>For each bug, we record all of the prompts, responses, and intermediate outputs.</p> <blockquote> <p>NEW: We have released the newest results for TSE'25 paper "SOAPFL: A Standard Operating Procedure for LLM-based Method-Level Fault Localization". The results can be found in the `<a href="https://zenodo.org/api/records/16938304/draft/files/SoapFL_results.zip/content" target="_blank" rel="noopener noreferrer">SoapFL_results.zip</a>` file!</p> </blockquote> <h3>Human Evaluation Results</h3> <p>The human evaluation results can be found in the file <code>AgentFL/EvaluationResult/DebugResult_d4j140_GPT35_human.xlsx</code></p> <h3>System Messages for Agents</h3> <ul> <li>Test Code Reviewer:</li> </ul> <blockquote> <p>You are a Test Code Reviewer. We share a common interest in collaborating to successfully locate the buggy code that cause the test suite to fail. You can examine the test code and the initialized classes to analyze the similar behavior of the failed tests within the test suite. To locate the bug, you must write a response that appropriately solves the requested instruction based on your expertise.</p> </blockquote> <ul> <li>Source Code Reviewer</li> </ul> <blockquote> <p>You are a Source Code Reviewer. we are both working at DebugDev. We share a common interest in collaborating to successfully locate the buggy code that cause the test suite to fail. Your main responsibilities is to generate a comment for each covered method base on the method call relationship. To locate the bug, you must write a response that appropriately solves the requested instruction based on your expertise.</p> </blockquote> <ul> <li>Software Test Engineer</li> </ul> <blockquote> <p>You are a Software Test Engineer. We share a common interest in collaborating to successfully locate the buggy code that cause the test suite to fail. You main responsibilities include examining the information of the failed tests to analyze the possible causes of the test failures, and determining the method that need to be fixed. To locate the bug, you must write a response that appropriately solves the requested instruction based on your expertise.</p> </blockquote> <ul> <li>Software Architect</li> </ul> <blockquote> <p>You are a Software Architect. We share a common interest in collaborating to successfully locate the buggy code that cause the test suite to fail. You are very familiar with the architecture of the software, the functions of each class and method in the software. You main responsibilities include examining the given information to locate the possible buggy classes and buggy methods. To locate the bug, you must write a response that appropriately solves the requested instruction based on your expertise.</p> </blockquote>
Respostas ao questionário Utilização de LLM no desenvolvimento de código de programação
<p>Respostas ao questionário "Utilização de LLM no desenvolvimento de código de programação"</p>
Replication Package for "Cracks in The Stack: Hidden Vulnerabilities and Licensing Risks in LLM Pre-Training Datasets"
<p>Replication Package for "Cracks in The Stack: Hidden Vulnerabilities and Licensing Risks in LLM Pre-Training Datasets"<br><br>Includes datasets, and bash scripts.</p>
Supplementary Material, Towards the LLM-Based Generation of Formal Specifications from Natural-Language Contracts: Early Experiments with Symboleo
<p>This repository contains all the files used in the experiments described in the paper "Towards the LLM-Based Generation of Formal Specifications from Natural-Language Contracts: Early Experiments with Symboleo", which appeared in "RAISE 2025: Requirements engineering for AI-powered SoftwarE", an ICSE 2025 workshop, Ottawa, May 3, 2025.</p>
Exploring Transparency Concerns in Social Media App Reviews: A Human-LLM Perspective
<p>This is a dataset for the "Exploring Transparency Concerns in Social Media App Reviews: A Human-LLM Perspective." paper submitted to EASE2025.</p>
Supplementary Materials for LAMeD: LLM-generated Annotations for Memory leak Detection
Open the record for dataset details and reuse information.
Artifact of An LLM-based Readability Measurement for Unit Tests' Context-aware Inputs
<div> <p><strong>Artifact of An LLM-based Readability Measurement for Unit Tests' Context-aware Inputs</strong></p> <p> </p> <p>A README.md file can be found in the base folder after unzipping the artifact.</p> <p> </p> </div>
AugmenTest: Enhancing Tests with LLM-driven Oracles - Replication package
<p>Replication package for the paper "AugmenTest: Enhancing Tests with LLM-driven Oracles"</p>
Artifacts for paper "CITYWALK: Enhancing LLM-Based C++ Unit Test Generation via Project-Dependency Awareness and Language-Specific Knowledge" submitted to TOSEM
<p>The project includes the data and code used in the submitted TOSEM paper titled "CITYWALK: Enhancing LLM-Based C++ Unit Test Generation via Project-Dependency Awareness and Language-Specific Knowledge"</p>
Using LLM Models and Explainable ML to Analyse Biomarkers at Single Cell Level
<p>Single-cell RNA sequencing (scRNA-seq) technology has significantly advanced our understanding of the diversity of cells and how this diversity is implicated in diseases. Yet, translating these findings across various scRNA-seq datasets poses challenges due to technical variability and dataset-specific biases. To overcome this, we present a novel approach that employs both an LLM-based framework and explainable machine learning to facilitate generalization across single-cell datasets and identify gene signatures to capture disease-driven transcriptional changes. Our approach uses scBERT, which harnesses shared transcriptomic features among cell types to establish consistent cell-type annotations across multiple scRNA-seq datasets. Additionally, we employ a symbolic regression algorithm to pinpoint highly relevant yet minimally redundant models and features for inferring a cell type’s disease state based on its transcriptomic profile. We ascertain the versatility of these cell-specific gene signatures across datasets, showcasing their resilience as molecular markers to pinpoint and characterize disease-associated cell types. Validation is carried out using four publicly available scRNA-seq datasets from both healthy individuals and those suffering from ulcerative colitis (UC). This demonstrates our approach’s efficacy in bridging disparities specific to different datasets, fostering comparative analyses. Notably, the simplicity and symbolic nature of the retrieved gene signatures facilitate their interpretability, allowing us to elucidate underlying molecular disease mechanisms using these models.</p>
Towards Human-interpretable Explanation in Code Clone Detection using LLM-based Post Hoc Explainer
<h2><strong>Overview</strong></h2> <div>This artifact accompanies the APSEC 2025 Research Track submission on <em>Towards Human-interpretable Explanation in Code Clone Detection using LLM-based Post Hoc Explainer</em>. The research focuses on using Large Language Models (LLMs) with local neighborhood sampling with in-context learning to provide explanations for GraphCodeBERT's code clone detection predictions.</div> <p> </p> <h2>Artifact Contents</h2> <h3>1. Core Implementation</h3> <div> <ul> <li> GraphCodeBERT_CloneDetection.ipynb: Complete Jupyter notebook containing: <ul> <li>GraphCodeBERT model setup and training</li> <li>Code clone detection experiments on BigCloneBench and Google Code Jam datasets</li> <li>In-Context Learning prompt generation for explanation</li> <li>LLM-based explanation pipeline</li> </ul> </li> </ul> </div> <h3>2. In-Context Learning Prompts (ICL_Prompts)</h3> <div>The ICL_Prompts directory contains structured prompts for 10 different Google Code Jam pairs (data_GoogleCodeJam_Pair1 through data_GoogleCodeJam_Pair10) used in the manual validation step, each with varying context lengths:</div> <ul> <li>layer_size4.txt: prompts with 4 examples</li> <li>layer_size8.txt: prompts with 8 examples</li> <li>layer_size16.txt: prompts with 16 examples</li> <li>layer_size32.txt: prompts with 32 examples</li> <li>layer_size48.txt: prompts with 48 examples</li> <li>layer_size64.txt: prompts with 64 examples</li> </ul> <div>Each prompt file is based on the prompt template structure:</div> <ul> <li><strong>Context:</strong> Description of GraphCodeBERT model</li> <li><strong>Dataset:</strong> Examples with code pairs, confidence scores, and predictions</li> <li><strong>Question:</strong> A query asking for explanation of a specific code pair</li> <li><strong>Instructions</strong> for generating explanations</li> </ul> <h3>3. Validation Results</h3> <div> <ul> <li>Manual_Validation_Results.xlsx: Manual evaluation results of generated explanations including quality assessments and human annotations.</li> </ul> </div> <h2>Usage Instructions</h2> <h3>Prerequisites</h3> <div><code>pip install torch transformers tree_sitter scikit-learn</code></div> <div> </div> <h3>Running the Code</h3> <ol> <li>Open `GraphCodeBERT_CloneDetection.ipynb` in Jupyter/Colab</li> <li>Follow the notebook cells to: <ol> <li>Set up the GraphCodeBERT model</li> <li>Load and preprocess datasets</li> <li>Run clone detection experiments</li> <li>Generate prompts for explanation</li> </ol> </li> </ol> <h3>File Organization</h3> <div><code>├── GraphCodeBERT_CloneDetection.ipynb # Main implementation</code></div> <div><code>├── Manual_Validation_Results.xlsx # Human evaluation results</code></div> <div><code>├── ICL_Prompts/ # Generated prompts</code></div> <div><code>│ ├── data_GoogleCodeJam_Pair1/ # Pair 1 prompts</code></div> <div><code>│ │ ├── layer_size4.txt # 4-example prompts</code></div> <div><code>│ │ ├── layer_size8.txt # 8-example prompts</code></div> <div><code>│ │ └── ... # Other sizes</code></div> <div><code>│ ├── data_GoogleCodeJam_Pair2/ # Pair 2 prompts</code></div> <div><code>│ └── ... # Additional pairs</code></div> <div><code>└── README.md # This file</code></div> <h3>Citation</h3> <div>Please cite this artefact using the DOI provided by Zenodo.</div> <h3>Contact</h3> <div>For questions about this artifact or research, please contact the authors [to be updated due to double-blind policy].</div>
Physician Response Evaluation With Contextual Insights vs. Standard Engines - Artificial Intelligence RAG vs LLM Clinical Decision Support
ClinicalTrials.gov study NCT07037940. IPD Sharing: NO. Countries: 1. Publications: 6.
Young People Aged 18-25 With Depression or Anxiety Mood Participate in an LLM-based Digital Dialogue Intervention Study
ClinicalTrials.gov study NCT06346496. IPD Sharing: Not stated. Countries: 1. Publications: 1.
Application of LLM Care and Related Affective Computing Systems on Persons With Special Needs
ClinicalTrials.gov study NCT04390321. IPD Sharing: NO. Countries: 1. Publications: 4.
LLM in Urodynamic Education
ClinicalTrials.gov study NCT07394335. IPD Sharing: NO. Countries: 1. Publications: 0.
ScienceDex guides
Understand access before you commit
These curated guides explain access requirements, typical timelines, costs, and reuse considerations for widely used research datasets.
Allen Brain Atlas
Allen Brain Atlas is an Allen Institute collection of brain map atlases, datasets, APIs, and analysis tools covering mouse, human, and non-human primate brain resources.
Annotated Behaviour and Observability Dataset (ABODe)
ABODe is a University of Edinburgh DataShare dataset for behavior classification in group-housed mice using home-cage video, identities, bounding boxes, ground-plate positions, and annotator labels.
DANDI Archive for NWB datasets
DANDI is a BRAIN Initiative archive for publishing and sharing neurophysiology data, including electrophysiology, optophysiology, and behavioral data packaged as NWB and related standards.
International Brain Laboratory public data
The International Brain Laboratory public data releases expose standardized mouse decision-making experiments, including Neuropixels recordings, widefield calcium imaging, behavior, and session metadata accessed through the ONE API.
OpenNeuro
OpenNeuro is a free, open platform for sharing neuroimaging datasets, with public search, dataset pages, and download paths for web, S3, DataLad, and the OpenNeuro CLI.