Update README.md

Prior Art
Anima Core: identity persistence with emotional weighting

Thane AI: self-authored context compaction

Qwen Episodic Summary: structured JSON session snapshots

Forge Protocol: topological emotional constraints

VividnessMem: neurochemically-inspired memory decay

Research Paper
Full concept paper with formal schema, differentiation analysis, and empirical validation protocol:

Chetan Sharma. Episodic Memory for AI Personas via Self-Authored Emotional State Snapshots: The yesterday.json Architecture. Zenodo, 15 May 2026.

DOI: 10.5281/zenodo.20191876

Blog: yesterday-json.blogspot.com

Citation
bibtex
@misc {sharma2026yesterdayjson,
author = {Chetan Sharma},
title = {Episodic Memory for AI Personas via Self-Authored Emotional State Snapshots: The yesterday.json Architecture},
year = {2026},
month = may,
doi = {10.5281/zenodo.20191876},
publisher = {Zenodo},
url = {https://zenodo.org/records/20191876}
}
Author
Chetan Sharma — Independent Researcher, Kolkata, India.

Blog: yesterday-json.blogspot.com

Zenodo: 10.5281/zenodo.20191876

License
This model card and the yesterday.json architecture concept are licensed under CC BY 4.0.

README.md

@@ -1,3 +1,67 @@
 ---
 license: cc-by-4.0
+pipeline_tag: text-generation
+tags:
+- ai-personas
+- digital-twins
+- episodic-memory
+- emotional-continuity
+- session-handoff
+- llm-agents
+- self-authored-memory
+- persona-continuity
+library_name: transformers
+language:
+- en
 ---
+
+# yesterday.json — Giving AI Personas Episodic Memory
+
+A lightweight (≤20 KB) episodic memory architecture where AI personas write their own emotional state snapshots at session end and read them at next startup, enabling emotional continuity across session boundaries.
+
+## Overview
+
+AI personas wake up blank. Every session is a cold start. Existing memory systems store *what happened* but systematically discard *how the interaction felt*. The yesterday.json architecture closes this gap.
+
+At the end of every session, the persona is prompted to write a brief, private note to its future self — a single JSON file (≤20 KB) containing:
+
+- **Emotional signature:** dominant mood, mood trajectory, emotional residue
+- **Open threads:** what was being discussed, current position, priority
+- **Decisions and insights:** conclusions reached, realizations
+- **Self-state snapshot:** current preoccupations, internal conflicts, mood susceptibility
+- **Last words:** a one-line handoff to tomorrow
+
+At the start of the next session, this snapshot is injected into the system prompt alongside the persona's permanent psychological constitution. The persona doesn't need a full transcript — it has *glimpses*, and its own cognitive architecture fills in the rest.
+
+## What Makes It Novel
+
+yesterday.json is the first system to combine six specific features:
+
+| Feature | Existing Work | yesterday.json |
+|--------|:---:|:---:|
+| Self-authored by the persona | Thane, Anima | ✓ |
+| JSON-structured schema | Qwen Episodic Summary | ✓ |
+| Emotional signature + mood trajectory | Forge (YAML), VividnessMem (weights) | ✓ |
+| Open thread tracking with priority | Qwen (tasks only) | ✓ |
+| Session-end automatic trigger | Qwen, Anima | ✓ |
+| Designed for digital twin identity continuity | — | ✓ |
+
+## Key Design Choices
+
+- **Self-authorship:** The persona writes its own memory — no external summarizer.
+- **Intentional rolling amnesia:** Only the most recent session is preserved, mimicking human short-term memory.
+- **Emotional carryover:** If yesterday ended in melancholy, tomorrow begins with its residue — not a blank slate.
+- **No model retraining required:** Works with any LLM, any framework, any runtime.
+
+## Minimal Implementation
+
+```python
+# At session start: inject yesterday's state
+yesterday_context = load_yesterday("persona_memory/yesterday.json")
+system_prompt = f"{PERSONA_CONSTITUTION}\n\n{yesterday_context}"
+
+# At session end: persona reflects and writes
+reflection_prompt = """The session is ending. Write a brief private note
+to your future self. Include what mattered emotionally, what remains
+unresolved, and what you were in the middle of. Output as JSON.
+Keep it under 20 KB."""