Research, software, data and methods. Inspect the context, compare your options, and take a useful next step.
Upstream project repository
Resampling methods and pipeline tools for learning from imbalanced classes.
Investigate How does the placement of resampling affect a fixed imbalanced-class evaluation?
Upstream project repository
Evaluation framework for composing tasks, model interactions, scoring and inspectable logs.
Investigate Does a deterministic mock-model evaluation preserve the input, scoring rule and retry history needed to audit every result?
UCI Machine Learning Repository
Radar-return descriptors for testing classification and calibration under limited sample sizes.
Investigate How does ionosphere classification calibration change across repeated fixed stratified splits?
Upstream project repository
Benchmark repository for evaluating language models on biology and laboratory research knowledge tasks.
Investigate Can a small non-procedural, rights-reviewed subset support an auditable scoring and contamination-risk review?
Upstream project repository
Task and model interfaces for reproducible language-model evaluation with inspectable prompting and scoring choices.
Investigate Can a tiny authored multiple-choice task expose the effect of length normalization on the selected answer?
UCI Machine Learning Repository
Leaf image descriptors for comparing small botanical classification baselines.
Investigate Which leaf classes are unstable under repeated fixed splits and train-only feature scaling?
ReScience C
A replication report on learning from labels that contain errors, with source metadata identifying a Python deep-learning implementation.
Investigate How does a prespecified label-corruption schedule affect held-out accuracy and calibration across fixed seeds?
UCI Machine Learning Repository
A synthetic display-recognition problem with controllable observation noise.
Investigate Does a classifier's error track the documented LED noise mechanism on a frozen evaluation design?
Upstream project repository
Language-model evaluation tooling for tracking prompt construction, scoring choices and evaluation configuration.
Investigate Can two declared scoring configurations be compared without changing task examples or model revision?
Upstream project repository
Model-provider interface and proxy tooling for inspecting request normalization, usage accounting and backend differences.
Investigate Can a local mock provider show whether retries and usage fields remain traceable without issuing any real model calls?
Upstream project repository
Code-model benchmark repository with time-aware task collections for studying evaluation provenance and data contamination.
Investigate Can task timestamps and release identifiers support a reproducible evaluation cut without silently mixing newer problems?
Upstream project repository
C/C++ language-model inference implementation for studying quantization and execution tradeoffs across supported hardware.
Investigate What output deviation and resource changes accompany one fixed quantization choice on an independently licensed tiny model?
Recorded license evidence is scoped to each source; it is not a blanket permission or a verification result.