dikwp

Purpose-aware AI agent benchmark: 72 synthetic scenarios, four reproducible baselines and inspectable permission traces. Python, offline.

AI-agent black-box recorder for tamper-evident DIKWP traces, incident replay, provenance and audit.

Claim-to-evidence verification for AI answers and retrieval-augmented generation: provenance, unsupported-claim detection and reversible review.

Offline functional white-box evaluation for LLMs across DIKWP transformations, inspectable artifacts, risks and boundaries.

DIKWP semantic-network prototype: 25 transformation classes, multi-observer conflicts and reproducible analysis. Python, offline.

Offline research studio for artificial-consciousness indicators, claim levels, evidence ledgers, scenarios and governance.

Compile mathematical problems into deterministic, hash-linked, machine-readable audit certificates with explicit proof boundaries.

Offline continuity and autonomous-learning research system for K-12 education under uncertain AI futures, with rights, agency, resilience and reversible pathways.

Yucong Duan's research hub: DIKWP graphs, artificial consciousness, semantic mathematics, auditable AI, publications and reproducible open-source experiments.

Offline opportunity planning: funded demand, DIKWP capability maps, market saturation, competing scenarios and reviewable evidence receipts. Python + browser; synthetic research examples.

Evidence-scoped transparent economy lab: bounded opportunity tests, exact contribution accounting, scenario analysis and replayable receipts. Offline Python and browser tools.

Evidence and context auditing for AI: inspect incentives, appeals and agent lineage, then validate outputs. Python and offline HTML.

Buyer-committed demand, shared-capacity planning and signed outcome records. Python, SQLite and an offline browser planner with reproducible synthetic tests.