An open library for your next question. Public pilot
Executable Science
Log inCreate account
THE OPEN LIBRARY

Find your next starting point.

Research, software, data and methods. Inspect the context, compare your options, and take a useful next step.

246 resources · Curated orderClear filters
Software

EvalPlus

Upstream project repository

Code-generation evaluation tooling and expanded testing resources for studying test coverage and result reproducibility.

Investigate Can a hand-authored incorrect solution pass a weak fixture but fail a separately specified edge-case test?

Listing reviewed · source terms apply
Software

FlashAttention

Upstream project repository

Attention-kernel implementations for comparing memory-efficient computation with a directly specified attention reference.

Investigate Does one supported attention kernel match dense scaled dot-product attention across mask and sequence-length edge cases?

Listing reviewed · source terms apply
Software

Flax

Upstream project repository

Neural-network tooling built around JAX for investigating explicit state, transformations and model composition.

Investigate Does a tiny stateful model preserve its declared parameter/state boundary through a training update and checkpoint?

Listing reviewed · source terms apply
Datasets

Forest Fires

UCI Machine Learning Repository

Weather and fire observations for studying skewed environmental regression targets.

Investigate Does log-target modeling improve forest-fire prediction on a frozen spatial-temporal evaluation?

Listing reviewed · source terms apply
Datasets

Forest type mapping

UCI Machine Learning Repository

Remote-sensing descriptors for examining forest-type identification across data groups.

Investigate How does forest-type accuracy change when location or acquisition groups are held out?

Listing reviewed · source terms apply
Datasets

Forty Soybean Cultivars from Subsequent Harvests

UCI Machine Learning Repository

Soybean cultivar observations across harvests for examining agricultural domain shift.

Investigate How does cultivar identification change when one harvest is withheld entirely?

Listing reviewed · source terms apply
Datasets

Gas Sensor Array Drift Dataset

UCI Machine Learning Repository

Gas-sensor responses over acquisition batches for investigating calibration drift.

Investigate How much does gas classification deteriorate when later acquisition batches are held out?

Listing reviewed · source terms apply
Datasets

Gas sensor array low-concentration

UCI Machine Learning Repository

Low-concentration gas-sensor observations for investigating detection limits and calibration stability.

Investigate How does low-concentration gas recognition change when acquisition batches are held out?

Listing reviewed · source terms apply
Datasets

Gas sensor array temperature modulation

UCI Machine Learning Repository

Temperature-modulated gas-sensor records for studying dynamic calibration and drift.

Investigate How does gas recognition change when temperature cycles or acquisition sessions are held out?

Listing reviewed · source terms apply
Datasets

Gas sensor array under flow modulation

UCI Machine Learning Repository

Gas-sensor measurements under flow modulation for studying dynamic chemical discrimination.

Investigate Can temporal summary features retain gas discrimination under acquisition-group holdout?

Listing reviewed · source terms apply
Datasets

Gas Turbine CO and NOx Emission Data Set

UCI Machine Learning Repository

Turbine operating variables and emission measurements for investigating physical prediction over time.

Investigate How does emissions prediction change when entire acquisition years are withheld?

Listing reviewed · source terms apply
Software

Gin Config

Upstream project repository

Configuration injection for Python functions and classes, useful for making experiment parameter binding explicit.

Investigate Can a small bound-function experiment reproduce the intended parameter values after configuration serialization?

Listing reviewed · source terms apply

Recorded license evidence is scoped to each source; it is not a blanket permission or a verification result.