K. Takahashi

Research Hub Overview

This site is the canonical entry point for K. Takahashi's research program on safe and auditable autonomous intelligence.

It is optimized for both human readers and automated systems through semantic HTML, Schema.org JSON-LD, DOI-linked publications, RSS updates, and crawlable navigation paths.

Core Research Areas

• Observable-only and no-meta governance for AI safety
• Auditable control and verification protocols for autonomous systems
• Information-theoretic foundations of persistent semantics and ontology drift
• Category-theoretic structure and formal foundations of intelligence
• First-principles formalization of intelligence using information theory, geometry, and controllability constraints
• Robust bottleneck diagnosis from public data via no-meta dynamic programming, anytime confidence sequences, and dynamic IQC
• Fail-closed verification pipelines with deterministic replay, canonical logs, and audit certificates
• Proof-carrying and evidence-carrying autonomy for reusable, verifiable AI/LLM workflows under compute and energy budgets

Machine-Readable Endpoints

• no-meta-observable-index.html: canonical YAML field guide for the no-meta / observable-only series
• works.html: full publication list with structured metadata and DOI links
• feed.xml: RSS feed for update polling
• robots.txt: crawler access policy
• sitemap.xml: crawl discovery map
• CITATION.cff: citation metadata for scholarly indexing

Additional Entry Points

• No-Meta / Observable-Only Series Index
• Self-Concealing Information / Observer-Modifying Dynamics
• Auditability / Replay / Fail-Closed AI Systems
• Long-Running Agents / Memory Limits / Multi-Agent Reasoning
• Provenance / Records / Semantic Interfaces
• Self-Improving Intelligence Foundations
• Theory Translation / Comparative Mathematics
• Propagation / Front Dynamics / Persistence

Collaboration

Scientific critique, replication attempts, and cross-disciplinary collaboration are welcome.

Research Support / Donations

This work is conducted as independent research to reduce institutional and personal bias. Support helps sustain research, publishing, and open repositories.

Accepted assets and networks are explicitly limited. Unsupported assets may be ignored and may not be recoverable. Do not send NFTs or unknown tokens. High-risk or compliance-sensitive funds may be held and may not be converted immediately.

Unsupported assets may be ignored and may not be recoverable. High-risk or compliance-sensitive funds may be held and may not be converted immediately.