🧠 KokoroSystem

A plug-and-play emotional framework for AI, enabled through revolutionary PDF-Prompting.Everything but the kitchen sink

PDF-Prompting → math-driven KRV (ER/GR/SR/IHR) → response generation. Not keyword-based; includes A/B verification and Raw vs Kokoro-processed comparison.

📄 Formula Flow PDF 📄 ILTV Framework (Minimal Configuration) ▶ Run the Kaggle Demo 🛠 Repo

TL;DR

1) What it does

Attaches an emotional layer to LLMs. Computes KRV via formulas and modulates responses transparently.

2) Why it’s built

Move beyond black-box/keyword “emotion AI” toward explainable, testable emotional cognition.

3) How to verify

Run the A/B test below: same words, different time/vector → different KRV & TR_Extended.

A/B Quick Check (3 lines)

A. I’m anxious about tomorrow’s talk.
B. Tomorrow’s talk went fine, but I still feel uneasy.

Watch: A → GR↑ (future-oriented), B → SR/IHR↑ (post-event unease); TR_Extended shifts. Proves math-driven behavior beyond keyword matching.

Overview

KokoroSystem formalizes emotion as structured objects (Integrity, Layer, Time, Vector) and computes a Kokoro Resonance Vector (ER/GR/SR/IHR). The 3-stage pipeline is: Llama2 emotion extraction → mathematical KRV → response generation. Safety is governed by PMC (coherence checks).