Replication & reproducibility
Trace published work to its sources and a useful next check.
Choose a useful starting point.
Begin with an exact published result and identify what a new contribution would test. A rerun, an independently written implementation and a changed-data extension answer different questions. Preserve the upstream authorship and version, and state the scope of your own contribution.
These external articles remain publications of their original venues. Library inclusion is not editorial acceptance or independent verification here. Inspect source permissions, unresolved inputs and the proposed comparison before deciding whether a reproduction can be bounded and useful.
Questions to work through.
Turn a published ML replication into a bounded next check
Which specific claim in an existing label-smoothing or noisy-label replication can be mapped to accessible code, data and an affordable evaluation?
Explain a score difference before blaming the model
Do evaluation harnesses agree on the same frozen predictions once normalization, aggregation and failure handling are made explicit?
Explore the sources.
All 246 resources →Iris
UCI Machine Learning Repository
Flower measurements for a compact, interpretable multiclass baseline and leakage audit.
Investigate How much do iris classification results change when preprocessing is restricted to training folds?
Hypothesis
Upstream project repository
Property-based testing tools that generate and simplify test inputs.
Investigate Can a fixed property-based test expose a specified floating-point overclaim within a bounded search protocol?
2D elastodynamic metamaterials
UCI Machine Learning Repository
Pixelated metamaterial unit-cell designs and band-gap locations and widths for investigating simulation surrogate reliability.
Investigate How does a band-gap surrogate generalize to held-out metamaterial design regions?
3W dataset
UCI Machine Learning Repository
Oil-production process signals for studying event detection across operating scenarios.
Investigate How does event detection change when complete wells or production episodes are held out?
[Re] A general model of hippocampal and dorsal striatal learning and decision making
ReScience C
A replication of a model combining hippocampal and dorsal-striatal learning for spatial decision making.
Investigate How does disabling each modeled learning component change performance on a prespecified navigation test?
[Re] A Multi-Functional Synthetic Gene Network
ReScience C
An Octave replication of a multifunctional synthetic gene-network model involving oscillatory and switching behavior.
Investigate Which simulated regime boundaries remain stable under solver refinement and small parameter perturbations?
[Re] A Neurodynamical Model for Working Memory
ReScience C
A replication of a neurodynamical working-memory model using recurrent or echo-state networks.
Investigate How does recall error vary with delay length under fixed input and noise schedules?
[Re] A Reservoir Computing Model of Reward-Modulated Motor Learning and Automaticity
ReScience C
A replication of reward-modulated motor learning and automaticity in a reservoir-computing model.
Investigate How do learning curves and post-training variability depend on reservoir initialization under a fixed reward schedule?
[Re] A simple rule for the evolution of cooperation on graphs and social networks
ReScience
A replication about the evolution of cooperation on graphs and social networks.
Investigate How do cooperation frequencies vary across evaluator-generated graph families under a fixed update rule?
[Re] Assortative matching and search
ReScience C
A Python replication of a search-and-matching model framed through game theory and markets.
Investigate How do matching statistics change under prespecified search-cost settings, and are terminal states stable under further updates?
[Re] Badder Seeds: Reproducing the Evaluation of Lexical Methods for Bias Measurement
ReScience C
A replication examining lexical methods for measuring bias and sensitivity to seed choices.
Investigate How much does the selected bias metric vary across prespecified seed-set perturbations compared with its sampling uncertainty?
[Re] Bandit Theory and Thompson Sampling-guided Directed Evolution for Sequence Optimization
ReScience C
A replication about Thompson sampling and bandit-guided sequence optimization.
Investigate How does Thompson sampling compare with uniform allocation on evaluator-generated finite reward landscapes at equal budgets?