Team Ai
Datasetpublic

tempresearch00/SCOPE-BENCH

SCOPE-BENCH: Scaffold-Cluster Out-Of-Distribution Performance Evaluation Benchmark SCOPE-BENCH is a rigorous out-of-distribution (OOD) benchmark for molecular property prediction. Unlike conventional scaffold splits, SCOPE-BENCH creates structurally disjoint source and target domains by clustering molecules based on physicochemical descriptors, blocking shortcut learning, and revealing true extrapolation abilities. 📊 Dataset Splits (As used in the NeurIPS 2026… See the full description on the dataset page: https://huggingface.co/datasets/tempresearch00/SCOPE-BENCH.

sourceHugging Facecc-by-nc-4.0updated 5mo agoView on Hugging Face
0likes83downloads
Dataset Card

SCOPE-BENCH: Scaffold-Cluster Out-Of-Distribution Performance Evaluation Benchmark

SCOPE-BENCH is a rigorous out-of-distribution (OOD) benchmark for molecular property prediction. Unlike conventional scaffold splits, SCOPE-BENCH creates structurally disjoint source and target domains by clustering molecules based on physicochemical descriptors, blocking shortcut learning, and revealing true extrapolation abilities.

📊 Dataset Splits (As used in the NeurIPS 2026 paper)

SplitFilesClustersSamplesPurpose
Source (training)a0.csv, a1.csv, a6.csv, a8.csv, a9.csv, a11.csv694,562Supervised pre‑training + multi‑source adaptation pool
Validationa10.csv118,326Hyperparameter tuning
Target (test)a2.csv, a3.csv, a4.csv, a5.csv, a7.csv519,894Strict OOD evaluation (zero‑shot extrapolation)
Fine‑grained tasksscaffold_datasets1/*.csv (each ≥200 samples)16variesIndependent zero‑shot evaluations (Table 2)

🔬 Molecular Properties

Each CSV file contains the following columns (based on the QM9 dataset):

ColumnDescriptionUnit
SMILESSimplified molecular input line entry system–
HOMOHighest Occupied Molecular Orbital energyeV
LUMOLowest Unoccupied Molecular Orbital energyeV
`GAP`HOMO–LUMO gap (LUMO – HOMO)eV
(other columns)Additional QM9 properties (e.g., dipole moment, polarizability, etc.)various

🗂️ File Structure

SCOPE-BENCH/ ├── a0.csv ... a11.csv # 12 cluster files ├── scaffolddatasets1/ # 16 independent target scaffolds ├── README.md ├── datasetinfo.json └── LICENSE