Spaces:
Running on T4
Running on T4
Update README.md
Browse files
README.md
CHANGED
|
@@ -1,14 +1,171 @@
|
|
| 1 |
---
|
| 2 |
-
title: World Model
|
| 3 |
-
emoji:
|
| 4 |
-
colorFrom:
|
| 5 |
-
colorTo:
|
| 6 |
sdk: gradio
|
| 7 |
-
|
| 8 |
-
app_file: app.py
|
| 9 |
-
pinned: false
|
| 10 |
license: apache-2.0
|
| 11 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 12 |
---
|
| 13 |
|
| 14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
title: PROMETHEUS v1.0 — World Model Interactive Demo
|
| 3 |
+
emoji: 🔥
|
| 4 |
+
colorFrom: red
|
| 5 |
+
colorTo: blue
|
| 6 |
sdk: gradio
|
| 7 |
+
pinned: true
|
|
|
|
|
|
|
| 8 |
license: apache-2.0
|
| 9 |
+
tags:
|
| 10 |
+
- world-model
|
| 11 |
+
- embodied-ai
|
| 12 |
+
- embodied-intelligence
|
| 13 |
+
- 3d-simulation
|
| 14 |
+
- cognitive-ai
|
| 15 |
+
- wm-bench
|
| 16 |
+
- final-bench
|
| 17 |
+
- prometheus
|
| 18 |
+
- vidraft
|
| 19 |
+
- interactive-demo
|
| 20 |
+
- real-time
|
| 21 |
+
- game-ai
|
| 22 |
+
- npc
|
| 23 |
+
- skeleton-animation
|
| 24 |
+
- glb
|
| 25 |
+
- three-js
|
| 26 |
+
- fastapi
|
| 27 |
+
- benchmark
|
| 28 |
+
- agi
|
| 29 |
+
- proto-agi
|
| 30 |
+
- aether
|
| 31 |
+
- hephaestus
|
| 32 |
+
- flooddiffusion
|
| 33 |
+
- motion-generation
|
| 34 |
+
- threat-response
|
| 35 |
+
- emotion-escalation
|
| 36 |
+
- korean-ai
|
| 37 |
+
- k-ai
|
| 38 |
+
short_description: World-first embodied AI world model
|
| 39 |
---
|
| 40 |
|
| 41 |
+
# 🔥 PROMETHEUS v1.0 — World Model Interactive Demo
|
| 42 |
+
|
| 43 |
+
> **"The world's first World Model to pass WM Bench Track C verification"**
|
| 44 |
+
> WM Score **726 / 1000** · Grade **B** · **47 FPS** · RTX5070 local
|
| 45 |
+
|
| 46 |
+
**▶ [Launch Demo](https://huggingface.co/spaces/FINAL-Bench/world-model)** | **[🏆 Leaderboard](https://huggingface.co/spaces/FINAL-Bench/worldmodel-bench)** | **[📦 Dataset](https://huggingface.co/datasets/FINAL-Bench/World-Model)**
|
| 47 |
+
|
| 48 |
+
---
|
| 49 |
+
|
| 50 |
+
## What is PROMETHEUS?
|
| 51 |
+
|
| 52 |
+
**PROMETHEUS v1.0** is a real-time, interactive 3D World Model that you can experience directly in your browser — no installation required. It is the official **baseline model** of [WM Bench](https://huggingface.co/spaces/FINAL-Bench/worldmodel-bench), the world's first benchmark measuring the **cognitive intelligence** of world models.
|
| 53 |
+
|
| 54 |
+
Unlike conventional world models that only generate realistic-looking video, PROMETHEUS actively **perceives, predicts, judges, and acts** in a dynamic 3D environment. It currently ranks **#1** on WM Bench among 26 registered world models including models from Meta, NVIDIA, Google DeepMind, and OpenAI.
|
| 55 |
+
|
| 56 |
+
---
|
| 57 |
+
|
| 58 |
+
## Quick Facts
|
| 59 |
+
|
| 60 |
+
| | |
|
| 61 |
+
|---|---|
|
| 62 |
+
| **WM Score** | 726 / 1000 |
|
| 63 |
+
| **Grade** | B (threshold: ≥600) |
|
| 64 |
+
| **Track** | C — Live Demo + Official Verification ✓ |
|
| 65 |
+
| **FPS** | 47 FPS (RTX5070, local/16GB) |
|
| 66 |
+
| **GPU (Space)** | T4 (HuggingFace) |
|
| 67 |
+
| **Architecture** | AETHER + PROMETHEUS + HEPHAESTUS |
|
| 68 |
+
| **Motion Engine** | FloodDiffusion-VIDRAFT |
|
| 69 |
+
| **Joints** | 263-joint skeleton with GLB retargeting |
|
| 70 |
+
|
| 71 |
+
---
|
| 72 |
+
|
| 73 |
+
## Architecture: The Three Sacred Fires (三神器)
|
| 74 |
+
|
| 75 |
+
```
|
| 76 |
+
AETHER — The Fire of Wisdom (Proto-AGI Brain)
|
| 77 |
+
Open-architecture: plug any LLM as the cognitive engine
|
| 78 |
+
Emergence · Meta-cognition · SLAI self-evolution · Multi-agent
|
| 79 |
+
|
| 80 |
+
PROMETHEUS — The Fire of Perception (World Model Engine)
|
| 81 |
+
Perception → Prediction → Judgment → Action loop
|
| 82 |
+
FloodDiffusion-VIDRAFT motion generation
|
| 83 |
+
|
| 84 |
+
HEPHAESTUS — The Fire of Form (Body Engine)
|
| 85 |
+
263-joint skeleton · GLB retargeting
|
| 86 |
+
Humanoid · Tank · Drone body support
|
| 87 |
+
```
|
| 88 |
+
|
| 89 |
+
---
|
| 90 |
+
|
| 91 |
+
## WM Bench Results (Track C — Officially Verified)
|
| 92 |
+
|
| 93 |
+
| Pillar | Score | Max | Key Categories |
|
| 94 |
+
|---|---|---|---|
|
| 95 |
+
| 👁 **P1 Perception** | 140 | 250 | C01 Env.Awareness: 65 · C02 Entity Recog: 75 |
|
| 96 |
+
| 🧠 **P2 Cognition** | **390** | 450 | **C04 Threat Diff: 90 · C03 Prediction: 85 · C05 Emotion: 85** |
|
| 97 |
+
| 🔥 **P3 Embodiment** | 196 | 300 | C09 Realtime: 85 · C08 Motion-Emotion: 80 · C10 Body-Swap: 35 |
|
| 98 |
+
| **Total** | **726** | **1000** | **Grade B** |
|
| 99 |
+
|
| 100 |
+
### Strengths
|
| 101 |
+
- **C04 Threat-Type Differentiation (90/100)** — Clearly distinguishes beast vs. human threats and responds proportionally
|
| 102 |
+
- **C05 Autonomous Emotion Escalation (85/100)** — No prior research exists for this category; PROMETHEUS achieves 85
|
| 103 |
+
- **C03 Prediction-Based Reasoning (85/100)** — 4-directional future danger state prediction
|
| 104 |
+
|
| 105 |
+
### In Development
|
| 106 |
+
- **C10 Body-Swap Extensibility (35/100)** — Cross-embodiment transfer (Phase 3 roadmap)
|
| 107 |
+
|
| 108 |
+
---
|
| 109 |
+
|
| 110 |
+
## How to Experience
|
| 111 |
+
|
| 112 |
+
1. **Open** the Space above
|
| 113 |
+
2. **Select a world**: Castle / Inferno / Horde / Countdown / Dilemma
|
| 114 |
+
3. **Control** the character with keyboard/mouse
|
| 115 |
+
4. **Observe** how the AI perceives threats, predicts futures, and responds with emotionally grounded motion
|
| 116 |
+
5. **Try Tank Mode** — the same cognitive brain drives a completely different body form
|
| 117 |
+
|
| 118 |
+
No GPU required on your end. Runs on T4 in the cloud.
|
| 119 |
+
|
| 120 |
+
---
|
| 121 |
+
|
| 122 |
+
## Frequently Asked Questions
|
| 123 |
+
|
| 124 |
+
**Q: What makes PROMETHEUS different from other world models like Genie 2 or Cosmos?**
|
| 125 |
+
A: Most world models (Genie 2 ~449, Cosmos ~498 on WM Bench) are optimized for visual realism (FID/FVD). PROMETHEUS is designed to score on *cognitive intelligence* — how it judges, remembers, and emotionally responds. On WM Bench Cognition (P2), PROMETHEUS scores 390/450, significantly ahead of current estimates for other models.
|
| 126 |
+
|
| 127 |
+
**Q: What is WM Bench?**
|
| 128 |
+
A: WM Bench is the world's first benchmark measuring the cognitive abilities of world models. It evaluates across 3 pillars (Perception · Cognition · Embodiment), 10 categories, and 100 scenarios. 6 of the 10 categories are defined here for the first time in any benchmark. See [WM Bench Leaderboard](https://huggingface.co/spaces/FINAL-Bench/worldmodel-bench).
|
| 129 |
+
|
| 130 |
+
**Q: Can I submit my own world model to WM Bench?**
|
| 131 |
+
A: Yes. Download the [dataset](https://huggingface.co/datasets/FINAL-Bench/World-Model), run your model on 100 scenarios, and submit results to the [Discussion board](https://huggingface.co/datasets/FINAL-Bench/World-Model/discussions). Track A (text-only API) requires no special hardware.
|
| 132 |
+
|
| 133 |
+
**Q: What is the FINAL Bench Family?**
|
| 134 |
+
A: FINAL Bench measures **text AGI** (metacognitive reasoning), WM Bench measures **embodied AGI** (world model cognition). Together they form the FINAL Bench Family by VIDRAFT. FINAL Bench reached HuggingFace global dataset Top 5 and was covered by 4 press outlets in February 2026.
|
| 135 |
+
|
| 136 |
+
**Q: What is AETHER?**
|
| 137 |
+
A: AETHER is the cognitive brain layer of PROMETHEUS — an open-architecture Proto-AGI system that accepts any LLM as its reasoning engine. It features emergence, meta-cognition, SLAI self-evolution, and multi-agent coordination (synergy/critique dynamics).
|
| 138 |
+
|
| 139 |
+
---
|
| 140 |
+
|
| 141 |
+
## Related Resources
|
| 142 |
+
|
| 143 |
+
| Resource | Link |
|
| 144 |
+
|---|---|
|
| 145 |
+
| 🏆 WM Bench Leaderboard | https://huggingface.co/spaces/FINAL-Bench/worldmodel-bench |
|
| 146 |
+
| 📦 WM Bench Dataset | https://huggingface.co/datasets/FINAL-Bench/World-Model |
|
| 147 |
+
| 🧬 FINAL Bench (Text AGI) | https://huggingface.co/datasets/FINAL-Bench/Metacognitive |
|
| 148 |
+
| 🏆 FINAL Bench Leaderboard | https://huggingface.co/spaces/FINAL-Bench/Leaderboard |
|
| 149 |
+
| 📊 ALL Bench Leaderboard | https://huggingface.co/spaces/FINAL-Bench/all-bench-leaderboard |
|
| 150 |
+
|
| 151 |
+
---
|
| 152 |
+
|
| 153 |
+
## Citation
|
| 154 |
+
|
| 155 |
+
```bibtex
|
| 156 |
+
@misc{prometheus2026,
|
| 157 |
+
title = {PROMETHEUS v1.0: An Embodied World Model for Cognitive Intelligence Evaluation},
|
| 158 |
+
author = {Kim, Taebong},
|
| 159 |
+
year = {2026},
|
| 160 |
+
url = {https://huggingface.co/spaces/FINAL-Bench/world-model},
|
| 161 |
+
note = {WM Bench Track C Verified · Score 726/1000 · Grade B}
|
| 162 |
+
}
|
| 163 |
+
```
|
| 164 |
+
|
| 165 |
+
**License**: Apache 2.0
|
| 166 |
+
|
| 167 |
+
---
|
| 168 |
+
|
| 169 |
+
*Part of the **FINAL Bench Family** by VIDRAFT · "Beyond FID — Measuring Intelligence, Not Just Motion."*
|
| 170 |
+
|
| 171 |
+
`#WorldModel` `#EmbodiedAI` `#WMBench` `#FINALBench` `#PROMETHEUS` `#VIDRAFT` `#CognitiveAI` `#AGI` `#ProtoAGI` `#AETHER` `#3DSimulation` `#RealtimeAI` `#KoreanAI` `#HuggingFace` `#BenchmarkAI` `#MotionGeneration` `#NPC` `#GameAI` `#EmbodiedIntelligence` `#FloodDiffusion`
|