Veritheia Documentation

An environment for inquiry - complete documentation

View the Project on GitHub cyharyanto/veritheia

Foundational Papers

This directory contains the research papers that inform Veritheia’s architecture and methodology.

Papers

1. LLAssist: Simple Tools for Automating Literature Review Using Large Language Models

File: 2407.13993v3.pdf
Authors: Christoforus Yoga Haryanto
Year: 2024
Reference: arXiv:2407.13993v3 [cs.DL]
DOI: https://doi.org/10.48550/arXiv.2407.13993
Note: 10 pages, 3 figures, 1 table, presented at the 51st International Conference on Computers and Industrial Engineering (CIE51), 11 Dec 2024
Subjects: Digital Libraries (cs.DL); Artificial Intelligence (cs.AI)
Latest version: v3, 20 Dec 2024

This paper introduces LLAssist, an open-source tool designed to streamline literature reviews by leveraging Large Language Models and NLP techniques. The methodologies for systematic screening and constrained composition in Veritheia are directly derived from the patterns established in LLAssist.

2. EdgePrompt: Engineering Guardrail Techniques for Offline LLMs in K-12 Educational Settings

File: 3701716.3717810.pdf
Authors: Riza Alaudin Syah, Christoforus Yoga Haryanto, Emily Lomempow, Krishna Malik, Irvan Putra
Year: 2025
Reference: WWW ‘25 Companion Proceedings, Pages 1635-1638
DOI: https://doi.org/10.1145/3701716.3717810
Published: 23 May 2025

This paper establishes the epistemic sovereignty principles and offline-first architecture that underpin Veritheia’s design. The guardrail techniques and local processing requirements directly influence the system’s use of local LLMs and rejection of cloud-first approaches.

3. Progress: A Post-AI Manifesto

File: 2408.13775v1.pdf
Authors: Christoforus Yoga Haryanto
Year: 2024
Reference: arXiv:2408.13775 [cs.CY]
DOI: https://doi.org/10.48550/arXiv.2408.13775
Note: work-in-progress
Subjects: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
Submitted: 25 Aug 2024

This manifesto establishes Veritheia’s foundational principle: AI systems must preserve human agency while amplifying intellectual capabilities. The emphasis on human-AI symbiosis directly informs Veritheia’s design as an instrument that supports rather than replaces intellectual work.

4. Cognitive Silicon: An Architectural Blueprint for Post-Industrial Computing Systems

File: 2504.16622v1.pdf
Authors: Christoforus Yoga Haryanto, Emily Lomempow
Year: 2025
Reference: arXiv:2504.16622 [cs.AI]
DOI: https://doi.org/10.48550/arXiv.2504.16622
Note: Working Paper, 37 pages, 1 figure, 5 tables
Subjects: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
Submitted: 23 Apr 2025

This paper provides Veritheia’s architectural vision: computing systems that embed moral coherence and preserve human identity through irreversible constraints. Veritheia implements these principles through journey-specific projection spaces where insights require three inseparable components—source materials, conceptual framework, and reasoning path. This architecture naturally incentivizes authentic knowledge transfer: copying outputs without the generative journey produces meaningless artifacts, ensuring intellectual sovereignty at the system level.

5. Contextualized AI for Cyber Defense: An Automated Survey Using LLMs

File: 2409.13524v1.pdf
Authors: Christoforus Yoga Haryanto, Anne Maria Elvira, Trung Duc Nguyen, Minh Hieu Vu, Yoshiano Hartanto, Emily Lomempow, Arathi Arakala
Year: 2024
Reference: 2024 17th International Conference on Security of Information and Networks (SIN), 02-04 December 2024
DOI: 10.1109/SIN63213.2024.10871242
Also: arXiv:2409.13524v1 [cs.CR] - https://doi.org/10.48550/arXiv.2409.13524
Note: 8 pages, 2 figures, 4 tables

This paper demonstrates two LLM-assisted literature survey methodologies that directly inform Veritheia’s approach to document analysis. The paper’s Method B (using Gemma 2:9b for screening and Claude 3.5 Sonnet for full-text analysis) provides the theoretical foundation for Veritheia’s systematic screening process. The emphasis on contextualized AI—systems that utilize proprietary and domain-specific knowledge—aligns with Veritheia’s journey-specific document projections.

Relevance to Veritheia

These five papers establish the theoretical foundation for Veritheia’s architecture as epistemic infrastructure for intellectual sovereignty.

Post-AI Manifesto and Cognitive Silicon provide the philosophical foundation: human-AI symbiosis that preserves agency while amplifying capability through systems that embed moral coherence via irreversible constraints. EdgePrompt establishes epistemic sovereignty through local-first processing and systematic cloud rejection. LLAssist demonstrates AI as measurement instrument within user frameworks rather than generic oracle. Contextualized AI reveals that AI achieves highest value within domain-specific, proprietary knowledge contexts.

Together, these papers resolve the fundamental tension in knowledge work: achieving scale without sacrificing authorship. Veritheia’s architecture transforms these insights into operational reality—an environment where formation happens through authorship, understanding emerges through engagement, and insights remain irreducibly personal while supporting collaborative intellectual work.