Update README.md

README.md

@@ -1,25 +1,36 @@
-Loko Release Gate Records
+---
+title: Loko AI
+emoji: 🔥
+colorFrom: gray
+colorTo: blue
+sdk: static
+pinned: true
+---
 
-This repository contains public Release Gate decision artifacts for pinned vision model revisions.
+# Loko Release Gate Records
 
-Each run documents:
-Model repository and exact revision
-Dataset repository and exact revision
-Declared release predicate
-Deterministic PASS/FAIL decision
-Receipt bundle and offline report
+Public Release Gate decision artifacts for pinned vision model revisions.
 
-The purpose is simple:
-To demonstrate that model releases can be governed by an explicit, machine-verifiable decision process.
+## What this is
 
-Run A — Baseline Repro (COCO)
+This Space serves as the organization card for LokoAI and describes the public release decision artifacts published in the `release-gate-records` dataset.
+
+Each run includes:
+- Model repository + exact revision
+- Dataset repository + exact revision
+- Declared release predicate
+- Deterministic PASS/FAIL decision
+- Receipt bundle + offline report
+
+## Runs
+
+### Run A — Baseline Repro (COCO)
 Pinned YOLO model evaluated on COCO val2017 to reproduce benchmark surface.
-This run demonstrates deterministic reproducibility and release decision recording.
 
-Run B — Production Gate (Low-Light, Class-Scoped)
-The same pinned YOLO revision evaluated under operational criteria using a low-light detection dataset.
-This run demonstrates deployment-grade gating for a specific class (e.g., person) under stricter production predicates.
+### Run B — Production Gate (Low-Light, class-scoped)
+Same pinned YOLO revision evaluated under operational criteria using a low-light detection dataset (class-scoped, e.g., `person`).
+
+## What this is not
 
-What This Is Not
-This repository does not claim model superiority or benchmark competition.
+This does not claim model superiority or benchmark competition.  
 It demonstrates auditable release governance.