datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
solidity_contract_sftsolidity-findings-v2
Solidity Findings Corpus (v2)
Canonical dataset of static analysis findings for smart contract security research.
Rows: 93194
Schema Version: 2
Corpus Digest: b6030d08a24775b81d90d5b8b51892061cc2b43e153207708b4e6236c61ea0d8
Last Updated: 2026-09-17T12:28:47.385053+00:00
solidity-audit-cot
solidity-audit-cot
Long-CoT audit traces for Solidity contracts, generated by Claude Opus 4.7 (adaptive thinking, xhigh effort) over the spec→contract corpus from the Qwopus3.6-27B-solidity training pipeline.
This dataset is the Stage 2 training corpus for the multi-stage Qwopus3.6-27B-solidity model — designed to teach long-form security reasoning (8-15 paragraph chain-of-thought) anchored to real Solidity contracts.
Why this dataset exists
Public Solidity audit… See the full description on the dataset page: https://huggingface.co/datasets/samscrack/solidity-audit-cot.solidity-findings
Solidity Findings Corpus (v2)
Canonical dataset of static analysis findings for smart contract security research.
Rows: 69228
Schema Version: 2
Corpus Digest: 895c60aee0839eb2d541b47a67a1238b470683a40b65ae4272bcd3a7bb99a241
Last Updated: 2026-09-17T03:17:37.108286+00:00
solidity-base-sft-v2marin-starcoderdata_soliditysolidity-cpt-top10-quality
Solidity CPT Top-10% Quality-Filtered Corpus
A curated, deduplicated corpus of 23,471 modern Solidity source files (~86M tokens) intended for continued-pretraining (CPT) of code LLMs on smart-contract code.
It's the top 10% slice (by composite quality score) of a larger raw corpus that combined:
ASSERT-KTH/DISL — 514 k unique deployed Solidity files, deduped at file level
30 hand-picked GitHub blue-chip protocols (OpenZeppelin, Uniswap v2/v3/v4, Aave v3, Compound, Morpho… See the full description on the dataset page: https://huggingface.co/datasets/samscrack/solidity-cpt-top10-quality.Solidity-Dataset-Cleanedx23-solidity-vulnerabilities-audit-findings
X23 Vulnerability Dataset
The X23 Vulnerability Dataset is a normalized corpus of public smart-contract security audit findings, focused mostly on Solidity and EVM-based protocols. It includes deduplicated vulnerability findings extracted from public audit reports published between May 2019 and April 2026, with normalized titles, summaries, descriptions, severity/status/category fields, affected files/functions, remediation text, source report metadata, and attribution links where… See the full description on the dataset page: https://huggingface.co/datasets/daveytea/x23-solidity-vulnerabilities-audit-findings.Slither-Audited-Solidity-QA
Dataset Card for "Simple-Solidity-Slither-Vulnerabilities"
More Information needed
solidity-rekt-dataset
Solidity REKT Dataset
Adversarial smart contract security pipeline training data for three specialized models:
Models & Splits
Model
File
Samples
Purpose
NGMI 🚫
ngmi_train.jsonl
27,616
Vulnerability classifier (type + severity)
MOG 👑
mog_train.jsonl
34,664
Security auditor (explain + patch)
REKT 💀
rekt_train.jsonl
2,411
Exploit PoC generator
Sources
oxdev/smart-contract-security-audit-v2 (48.4K rows) — Code4rena, Sherlock… See the full description on the dataset page: https://huggingface.co/datasets/aissacas/solidity-rekt-dataset.solidity-base
Solidity Dataset with Prompts
The "Solidity Dataset with Prompts" (braindao/Solidity-Dataset) is an enhanced version of the original seyyedaliayati/solidity-dataset. This dataset includes not only pure Solidity code but also carefully crafted prompts to aid in various NLP tasks. These prompts facilitate:
Code Understanding: Enhancing models' ability to comprehend and explain Solidity code.
Code Generation: Encouraging models to generate specific Solidity code snippets.
Error… See the full description on the dataset page: https://huggingface.co/datasets/braindao/solidity-base.solidity_errors_and_vulnerabilities
Solidity Vulnerabilities Dataset
📖 Overview
This dataset contains examples of common vulnerabilities in Solidity smart contracts, structured for use in Retrieval-Augmented Generation (RAG) systems.
It is intended to give LLMs context for:
Detecting vulnerabilities in Solidity code
Explaining security issues in simple terms
Suggesting fixes and mitigations
Assessing the severity of the issue
🗂 Data Format
Each entry is a JSON object with the… See the full description on the dataset page: https://huggingface.co/datasets/greatestyapper/solidity_errors_and_vulnerabilities.solidity-security-findings
Solidity Security Findings Dataset
9,359 smart contract security findings from Code4rena audits.
Statistics
HIGH severity: 4,418 findings
MEDIUM severity: 4,910 findings
Sources: Code4rena, Immunefi
Use Cases
Training security audit models
Vulnerability classification
Smart contract analysis
reentrancy_solidity_function
Dataset Card for "reentrancy_solidity_function"
More Information needed
solidity_vulnerability_audit_dataset
Solidity Vulnerability Audit Dataset
Organization: gitmate AI
Dataset Summary
The Solidity Vulnerability Audit Dataset is a curated collection of Solidity smart contract code snippets paired with expert-written vulnerability audits. Each entry presents a real or realistic smart contract scenario, and the corresponding analysis identifies security vulnerabilities or confirms secure patterns. The dataset is designed for instruction-tuned large language models (LLMs) to… See the full description on the dataset page: https://huggingface.co/datasets/GitmateAI/solidity_vulnerability_audit_dataset.multi_class_solidity_function_vulnerabilty
Dataset Card for "multi_class_solidity_function_vulnerabilty"
More Information needed
solidity-base-sft-300k-v1soliditysolidity-vulnerability-eval-datasetsolidity-bettergpt-base-v2-promptssolidity-eval-2026
Solidity Eval (2026)
Agentic Solidity benchmark. Each task hands the agent a Foundry workspace where one function body in a real Etherscan-verified contract has been replaced with revert("TODO");. The agent edits, builds (forge build), and tests (forge test) inside a sandbox until it returns. Reward is the differential-fuzz pass rate (Diffusc + Echidna) of the model's body against the ground-truth body.
This dataset is intended for use with the hermes-agent Solidity Eval environment… See the full description on the dataset page: https://huggingface.co/datasets/samscrack/solidity-eval-2026.solidity-base-sft-100k-v1starcoder_solidity_finetunesolidity-bettergpt-base-v1w3sa-solidity-access-control
W3SA - Solidity Access Control Benchmark
This benchmark includes 18 test High Severity Access Control vulnerabilities, derived from real world contracts audited through Code4rena competitions. Credits to Zhuo for the initial data scraping and curation. We add a python wrapper and standardized evaluation framework for the evaluation of AI Models.
Project Statistics
These are from 12 different projects, and only the files containing the bug are included in this benchmark.… See the full description on the dataset page: https://huggingface.co/datasets/almanax/w3sa-solidity-access-control.solidity-base-sft-v3solidity-base-enrichedSolidity-contract-Specification-dataset-updatedsolidity-vuln-expert-reentrancy