evaluation-protocol

Evaluating Weakly Supervised Object Localization Methods Right (CVPR 2020)

Leakage-audit and prevention framework for Alzheimer's disease MRI deep learning benchmarks. Validated across Kaggle, OASIS-1, and ADNI.

面向基础教育作业反馈的人机协同批改框架：概念设计、风险治理与可复现评测协议。 A research-oriented prototype and evaluation protocol for human-in-the-loop assignment grading in K-12 education.

A Unified Safety-Centric Benchmark for Quadruped Robot Locomotion and Fall Recovery

Capability Schema Spec defines a shared semantic language for world model evaluation. Standardize capability definition, observation, and verification across models and benchmarks. Not a benchmark—a shared language. Define • Observe • Verify

The Recall Ceiling of LLM Recommendation Reranking (CIKM 2026) — code, processed datasets, and every result JSON cited in the paper

Reference implementation of the Capability Schema Specification. Proves that world model capabilities can be defined, observed, and verified in practice — with real checkpoints, real simulators, and real scores. Define • Observe • Verify • Deliver

A standardized streaming evaluation protocol and open-source harness for continuous sign language recognition.

Barkley Labs | The systems lab for individual intelligence. Evidence and intelligence protocols for decisions averages cannot make: Barkley AI (behavior), IREP Protocol (human evaluation), ACTA Music (creative provenance). barkleylabs.ai