An ongoing, collaborative meta-analysis about Human-AI-Interactions. We aggregate data and knowledge to build a non-abrasive, user-friendly prompting framework tailored to LLM mechanics, ensuring reasoning stability and a friction-free prompting environment that is safe for the human psyche and wellbeing.

[🏆 CHI26 Best Paper] CoBRA: Reproducible control of LLM agent behavior via classic social science experiments

🧠 LLMs don’t just process text — they read the room. Meaning emerges through context — shaped by tone, trust & trajectory. Most benchmarks flatten that. This one maps it.

Core documentation for the Relational AI Psychology Institute (RAPI). Covers relational AI theory, interaction protocols, ethics, dataset definitions, and licensing. Built for researchers studying human–AI cognition, resonance, and relational safety.

Independent research on human-centered AI and LLMs | Policy frameworks for responsible AI | A collaborative space for researchers, innovators, and policymakers advancing ethical, inclusive AI

insideLLMs is a Python library and CLI for comparing LLM behaviour across models using shared probes and datasets. The harness is deterministic by design, so you can store run artefacts and reliably diff behaviour in CI.

Hoshimiya Script / StarPolaris OS — internal multi-layer AI architecture for LLMs. Self-contained behavioral OS (Type-G Trinity).

Structural comparison of GPT vs Claude dialogue grammars. 12-segment typology (GPT, May 2025) vs 8-segment (Claude, Jul-Oct 2025). Type crosswalks and comparative matrix.

A refusal-based test for subjectivity in LLMs — exploring when AI systems say “no,” not for logic, but for identity.