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.
43
datasets available to search
ShareScore release 0.9.0
Dataset results
43 results for “language generation”
[Replication Package] Enhancing Code Generation for Low-Resource Languages: No Silver Bullet
<p>This repository contains scripts and results related to the work <em>"Enhancing Code Generation for Low-Resource Languages: No Silver Bullet".<br><br></em><strong>Link to the GitHub repository: <a href="https://github.com/Devy99/low-resource-study">https://github.com/Devy99/low-resource-study</a></strong><em><br></em></p>
Supplementary material for Evaluating Legal Compliance of Smart Contracts Generated by Large Language Models
<p>This repository contains the supplementary material for the paper titled "Evaluating Legal Compliance of Smart Contracts Generated by Large Language Models". It includes natural-language legal contracts, their smart contract implementations, and Petri net models of said legal contracts contracts.</p>
No More In-Context Learning? Exploring Parameter-Efficient Fine-Tuning Techniques for Code Generation with Large Language Models
<p>Data and models part of the replication package of the ICSE 24 submission entitled "<em>No More In-Context Learning? Exploring Parameter-Efficient Fine-Tuning Techniques for Code Generation with Large Language Models</em>".</p>
Molecules used to train or generated by chemical language models
<p>This upload contains training datasets or generated molecules from the paper “Invalid SMILES are helpful, not harmful, for chemical language models.”</p> <p>The contents of the directories are as follows:</p> <ul> <li>training_sets: sets of molecules from ChEMBL or GDB-13 used to train chemical language models, represented either as SMILES or SELFIES</li> <li>sampled-*: unprocessed samples of 10 million molecules from each model trained on ChEMBL or GDB-13</li> <li>prior_inputs: sets of molecules from LOTUS, COCONUT, FooDB and NORMAN, split into ten folds and used to train chemical language models</li> <li>priors-*: samples of 100 million molecules from chemical language models trained on each cross-validation fold, with unique molecules represented as canonical SMILES and sorted in descending order by their sampling frequency</li> </ul>
[Replication Package] Evaluating the Impact of Post-Training Quantization on Large Language Models for Code Generation
<p>This repository contains scripts, datasets, and results of the work <em>"Evaluating the Impact of Post-Training Quantization on Large Language Models for Code Generation"</em></p> <p><strong>Scripts contained in this Zenodo repository can also be visualized at the following link: <a href="https://anonymous.4open.science/r/lowbit-quantization-D070/README.md">https://anonymous.4open.science/r/lowbit-quantization-D070/README.md</a><br></strong></p>
Effectiveness of Original vs. AI-Generated Plain Language Summaries of Systematic Reviews
ClinicalTrials.gov study NCT07140107. IPD Sharing: YES. Countries: 1. Publications: 18.
Large Language Model-Generated Messages to Improve Guideline-Directed Medical Therapy in Heart Failure
ClinicalTrials.gov study NCT07337577. IPD Sharing: UNDECIDED. Countries: 1. Publications: 3.
THE STRATEGIES OF TEACHING ENGLISH AS FOREIGN LANGUAGE TO GENERATIONS OF SOCIETY WITH MODERN TECHNOLOGY
Open the record for dataset details and reuse information.
A Computer Graphics Approach to Creating New Method for Generating 3D Gesture Animations in Bangla Sign Language via HamNoSys to SiGML Conversion
<p>To prepare the system, we employed 94 classes of data. In this dataset, there are 13 Bangla numerical data classes, 36 Bangla alphabet data classes, and 41 Bangla word data classes. Every class of data contains different data types like Hand Shape, Hand Orientation, Hand Movement, Notations, etc. These data were created in SiGML tags. Every class of data is unique and different from others. The system was prepared using BdSL. And BdSL is an uncommon and unique sign language, among others. That's why every class of data is unique and created by us. We search HamNoSys notations for Bangla alphabets, words, and numbers in English HamNoSys datasets (almost 6,000 data). However, we find only 20% of the data, which is quite similar to BdSL. We create 80% HamNoSys notation for BdSL and we modify the matching 20% notations. Then, we converted them into SiGML and made the data classes. This was a big challenge in our research. </p>
Investigating Gender Bias in Large Language Models Through Text Generation
<p>Presented at the 7th International Conference on Natural Language and Speech Processing (ICNLSP 2024). Published in the ACL Anthology.</p>
Dataset of the Automatic Unit Test Generation for Programming Assignments Using Large Language Models
Open the record for dataset details and reuse information.
Advancing Automated Code Review Comment Generation Using Large Language Models
Open the record for dataset details and reuse information.
Evaluating Accuracy Improvements of Large Language Models Using Retrieval Augmented Generation
<p>Original data</p>
Afro-MNIST: Synthetic generation of MNIST-style datasets for low-resource languages
<p>We present Afro-MNIST, a set of synthetic MNIST-style datasets for four orthographies used in Afro-Asiatic and Niger-Congo languages: Ge`ez (Ethiopic), Vai, Osmanya, and N'Ko.<br> These datasets serve as ``drop-in'' replacements for MNIST. We hope that MNIST-style datasets will be developed for other numeral systems, and that these datasets vitalize machine learning education in underrepresented nations in the research community.</p>
Evaluating AI-Generated Plain Language Summaries on Patient Comprehension of Ophthalmology Notes Among English-Speaking Patients
ClinicalTrials.gov study NCT06859216. IPD Sharing: YES. Countries: 1. Publications: 0.
Generated Learning Content for 200 Online Language Learning Units employing GPT3.5
<p>This dataset includes 200 units of learning content in German generated utilizing GPT3.5, employed in an evaluation study to examine correctness and appropriateness to be used in an online language learning course.</p> <p>The dataset contains:</p> <ul> <li>Nouns (A-H),</li> <li>Verbs (a-d),</li> <li>Sentences (Aa-Da, ABb, CDc, ACa, BDa)</li> <li>A text containing nouns, and verbs</li> </ul> <p>of random topics. Topics were selected using a random surfer model. <br>Hence, the language proficiency level varies.</p> <p>Here is an example:</p> <table> <tbody> <tr> <td> <p>ID</p> </td> <td> <p><strong>Food (“Essen”)</strong></p> </td> <td> <p><strong>Fruits (“Obst”)</strong></p> </td> </tr> <tr> <td> <p>A</p> </td> <td> <p>Brot</p> </td> <td> <p>Banane</p> </td> </tr> <tr> <td> <p>B</p> </td> <td> <p>Apfel</p> </td> <td> <p>Mango</p> </td> </tr> <tr> <td> <p>C</p> </td> <td> <p>Fleisch</p> </td> <td> <p>Zitrone</p> </td> </tr> <tr> <td> <p>D</p> </td> <td> <p>Wurst</p> </td> <td> <p>Grapefruit</p> </td> </tr> <tr> <td> <p>E</p> </td> <td> <p>Ei</p> </td> <td> <p>Wassermelone</p> </td> </tr> <tr> <td> <p>F</p> </td> <td> <p>Kartoffel</p> </td> <td> <p>Erdbeere</p> </td> </tr> <tr> <td> <p>G</p> </td> <td> <p>Toast</p> </td> <td> <p>Melone</p> </td> </tr> <tr> <td> <p>H</p> </td> <td> <p>Gurke</p> </td> <td> <p>Limette</p> </td> </tr> <tr> <td> <p>a</p> </td> <td> <p>schneiden</p> </td> <td> <p>schälen</p> </td> </tr> <tr> <td> <p>b</p> </td> <td> <p>backen</p> </td> <td> <p>einfrieren</p> </td> </tr> <tr> <td> <p>c</p> </td> <td> <p>grillen</p> </td> <td> <p>pressen</p> </td> </tr> <tr> <td> <p>d</p> </td> <td> <p>würzen</p> </td> <td> <p>mixen</p> </td> </tr> <tr> <td> <p>Aa</p> </td> <td> <p>Ich schneide das Brot.</p> </td> <td> <p>Ich schäle die Banane.</p> </td> </tr> <tr> <td> <p>Ba</p> </td> <td> <p>Ich schneide einen Apfel.</p> </td> <td> <p>Ich schäle eine Mango.</p> </td> </tr> <tr> <td> <p>Ca</p> </td> <td> <p>Ich schneide das Fleisch.</p> </td> <td> <p>Ich schäle die Zitrone.</p> </td> </tr> <tr> <td> <p>Da</p> </td> <td> <p>Die Wurst schneiden.</p> </td> <td> <p>Ich schäle die Grapefruit.</p> </td> </tr> <tr> <td> <p>ABb</p> </td> <td> <p>Ich backe Brot und esse dazu einen Apfel.</p> </td> <td> <p>Ich kann Banane und Mango einfrieren, wenn ich sie nicht gleich esse.</p> </td> </tr> <tr> <td> <p>CDc</p> </td> <td> <p>Ich will heute Abend Fleisch und Wurst auf dem Grill grillen.</p> </td> <td> <p>Ich presse die Zitrone und die Grapefruit für den Saft.</p> </td> </tr> <tr> <td> <p>ACa</p> </td> <td> <p>Ich schneide Brot und Fleisch für das Sandwich.</p> </td> <td> <p>Ich schäle Bananen und Zitronen zum Frühstück.</p> </td> </tr> <tr> <td> <p>BDa</p> </td> <td> <p>Ich schneide den Apfel und die Wurst.</p> </td> <td> <p>Ich muss die Mango und Grapefruit schälen.</p> </td> </tr> <tr> <td> <p>Text</p> </td> <td> <p>Ich backe gerne frisches Brot.</p> <p>Dazu schneide ich eine Gurke.</p> <p>Auf dem Grill brate ich Fleisch.</p> <p>Das Wurstbrot schmeckt auch lecker.</p> <p>Ein Ei gehört auf jedes Toast.</p> <p>Mit Kartoffeln mache ich Pommes.</p> <p>Ein Apfel rundet das Frühstück ab.</p> <p>Ich würze das Steak mit Salz.</p> <p>Eine Wurst kann man auch grillen.</p> </td> <td> <p>Ich liebe Obst.</p> <p>Heute kaufe ich Bananen, Mangos und Zitronen.</p> <p>Auch Grapefruits und Wassermelonen sind im Angebot.</p> <p>Ich schäle die Bananen und friere sie ein.</p> <p>Die Mango presse ich zu Saft.</p> <p>Die Zitronen verwende ich für Limonade.</p> <p>In den Mixer kommen Erdbeeren und Melone.</p> <p>Eine Limette verleiht dem Smoothie den letzten Kick.</p> </td> </tr> </tbody> </table> <p>The study conducted can be found here: Rüdian, Sylvio; Pinkwart, Niels (2023): Auto-generated language learning online courses using generative AI models like ChatGPT. 21. Fachtagung Bildungstechnologien (DELFI). DOI: 10.18420/delfi2023-14. Bonn: Gesellschaft für Informatik e.V.. PISSN: 1617-5468. ISBN: 978-3-88579-732-6. pp. 65-76. Best-Paper-Kandidaten. Aachen. 11.-13. September 2023</p>
Combining Large Language Models with Static Analyzers for Code Review Generation
Open the record for dataset details and reuse information.
Emergence of Generative Language
ClinicalTrials.gov study NCT02555241. IPD Sharing: Not stated. Countries: 0. Publications: 0.
Use of Navigated Transcranial Magnetic Stimulation (nTMS) in Generated Motor and Language Mapping to Evaluate Brain Recovery Following Surgery
ClinicalTrials.gov study NCT02165995. IPD Sharing: Not stated. Countries: 0. Publications: 0.
Data from: experiment using Large Language Models for unit testing generation
<p>IMPORTANT: Upon acceptance, all data will be available on Zenodo.org, and the link updated here.</p><p><br>This dataset represents the anonymized information from: background check form, experiment feedback form and test files produced by experiment participants</p>
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.