Spaces:
Running
Running
<html> | |
<head> | |
<meta charset="utf-8"> | |
<meta name="description" | |
content="MPCA is a novel, bio-inspired AI framework that moves beyond traditional machine learning models."> | |
<meta name="keywords" content="MPCA, MycoPhysarum, Cognitive Architecture, Bio-inspired AI, Slime Mold, Mycelium, Graph AI, Efficient AI"> | |
<meta name="viewport" content="width=device-width, initial-scale=1"> | |
<title>MycoPhysarum Cognitive Architecture (MPCA)</title> | |
<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" | |
rel="stylesheet"> | |
<link rel="stylesheet" href="./static/css/bulma.min.css"> | |
<link rel="stylesheet" href="./static/css/bulma-carousel.min.css"> | |
<link rel="stylesheet" href="./static/css/bulma-slider.min.css"> | |
<link rel="stylesheet" href="./static/css/fontawesome.all.min.css"> | |
<link rel="stylesheet" | |
href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css"> | |
<link rel="stylesheet" href="./static/css/index.css"> | |
<link rel="icon" href="./static/images/favicon.svg"> | |
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script> | |
<script defer src="./static/js/fontawesome.all.min.js"></script> | |
<script src="./static/js/bulma-carousel.min.js"></script> | |
<script src="./static/js/bulma-slider.min.js"></script> | |
<script src="./static/js/index.js"></script> | |
</head> | |
<body> | |
<section class="hero"> | |
<div class="hero-body"> | |
<div class="container is-max-desktop"> | |
<div class="columns is-centered"> | |
<div class="column has-text-centered"> | |
<h1 class="title is-1 publication-title">MycoPhysarum Cognitive Architecture (MPCA)</h1> | |
<div class="is-size-5 publication-authors"> | |
<span class="author-block"> | |
A Novel, Bio-Inspired AI Framework | |
</span> | |
</div> | |
<div class="is-size-5 publication-authors"> | |
<span class="author-block">Inspired by the decentralized intelligence of slime molds and mycorrhizal networks.</span> | |
</div> | |
<div class="column has-text-centered"> | |
<div class="publication-links"> | |
<!-- Code Link. --> | |
<span class="link-block"> | |
<a href="https://github.com/bahira/myco-physarum" target="_blank" | |
class="external-link button is-normal is-rounded is-dark"> | |
<span class="icon"> | |
<i class="fab fa-github"></i> | |
</span> | |
<span>Code</span> | |
</a> | |
</span> | |
</div> | |
</div> | |
</div> | |
</div> | |
</div> | |
</div> | |
</section> | |
<section class="hero teaser"> | |
<div class="container is-max-desktop"> | |
<div class="hero-body"> | |
<!-- Replaced video with a conceptual diagram --> | |
<img src="https://huggingface.co/bahira/myco-physarum/resolve/main/mpca.png" alt="MPCA Architecture Diagram" style="width: 100%; max-width: 800px; margin: auto; display: block; border-radius: 10px;"> | |
<h2 class="subtitle has-text-centered" style="margin-top: 2rem;"> | |
MPCA represents knowledge not as static data, but as a dynamic, living graph that evolves through interaction and self-reflection. | |
</h2> | |
</div> | |
</div> | |
</section> | |
<section class="hero is-light is-small"> | |
<div class="hero-body"> | |
<div class="container"> | |
<div id="results-carousel" class="carousel results-carousel"> | |
<div class="item"> | |
<div class="content" style="padding: 2rem;"> | |
<h3 class="title is-4">Mycelium</h3> | |
<p>The heart of the system. A <code>networkx</code> directed graph where nodes are concepts and edges represent the relationships between them. The strength of these connections is dynamic, changing based on usage and learning.</p> | |
</div> | |
</div> | |
<div class="item"> | |
<div class="content" style="padding: 2rem;"> | |
<h3 class="title is-4">Builder</h3> | |
<p>The architect of the Mycelium. The Builder ingests raw text, uses <code>spaCy</code>'s dependency parser to understand grammatical structure, and translates it into a rich graph of nodes and relationships.</p> | |
</div> | |
</div> | |
<div class="item"> | |
<div class="content" style="padding: 2rem;"> | |
<h3 class="title is-4">Solver</h3> | |
<p>The "consciousness" of MPCA. It traverses the Mycelium to find relevant pathways to answer queries. Implements Hebbian learning to reinforce successful paths and "dreaming" to form new speculative connections.</p> | |
</div> | |
</div> | |
<div class="item"> | |
<div class="content" style="padding: 2rem;"> | |
<h3 class="title is-4">Node</h3> | |
<p>The fundamental unit of knowledge. Each node represents a word and has a <code>type</code> (concept, action, property) and a <code>strength</code>, indicating its importance in the network.</p> | |
</div> | |
</div> | |
<div class="item"> | |
<div class="content" style="padding: 2rem;"> | |
<h3 class="title is-4">Spore</h3> | |
<p>A highly efficient persistence mechanism. A "spore" is a serialized (<code>pickle</code>) and compressed snapshot of the Mycelium, allowing the system's learned state to be saved and loaded from a tiny file (~9 MB).</p> | |
</div> | |
</div> | |
</div> | |
</div> | |
</div> | |
</section> | |
<section class="section"> | |
<div class="container is-max-desktop"> | |
<!-- Abstract --> | |
<div class="columns is-centered has-text-centered"> | |
<div class="column is-four-fifths"> | |
<h2 class="title is-3">Core Philosophy</h2> | |
<div class="content has-text-justified"> | |
<p> | |
The core of MPCA is the "Cognitive Mycelium," a graph-based knowledge structure. Unlike rigid, pre-trained models, the Mycelium is built from the ground up to understand the grammatical and conceptual relationships in language. It learns, forgets, and even "dreams" to form new connections, creating a resilient and emergent form of intelligence with a fraction of the computational overhead of conventional architectures. | |
</p> | |
<h2 class="title is-3" style="margin-top: 2rem;">Radical Efficiency: The Spore Advantage</h2> | |
<p> | |
A key breakthrough of the MPCA is its incredible efficiency. A fully-formed Cognitive Mycelium, built from a large dataset (~100k entries, >2.9M sentences), can be compressed into a <code>spore</code> file of only <strong>~9 MB</strong>. This stands in stark contrast to conventional AI models like Transformers (GPT-2 is ~500 MB; modern models are many gigabytes). | |
</p> | |
<p> | |
This efficiency is a direct result of the architecture's design. Instead of storing billions of statistical weights to predict tokens, the Mycelium stores a compressed graph of concepts and their relationships. It captures knowledge, not just statistical patterns, leading to a powerful, lightweight, and truly novel form of intelligence. | |
</p> | |
</div> | |
</div> | |
</div> | |
<!--/ Abstract. --> | |
<!-- Paper video. --> | |
<div class="columns is-centered has-text-centered"> | |
<div class="column is-four-fifths"> | |
<h2 class="title is-3">How to Run</h2> | |
<div class="content has-text-justified"> | |
<p><strong>1. Install Dependencies:</strong></p> | |
<pre><code>pip install -r requirements.txt | |
python -m spacy download en_core_web_sm</code></pre> | |
<p><strong>2. Build a New Mycelium:</strong></p> | |
<p>To create a new knowledge graph from a dataset, run the main script with the <code>--build</code> flag. The default dataset is <code>mlabonne/FineTome-100k</code>.</p> | |
<pre><code>python main.py --build --spore-file mycelium_new.spore --limit 1000</code></pre> | |
<p><strong>3. Interact with an Existing Mycelium:</strong></p> | |
<p>To chat with a pre-built Mycelium, use the <code>--interactive</code> flag.</p> | |
<pre><code>python main.py --interactive --spore-file mycelium.spore</code></pre> | |
</div> | |
</div> | |
</div> | |
<!--/ Paper video. --> | |
</div> | |
</section> | |
<section class="section"> | |
<div class="container is-max-desktop"> | |
<div class="columns is-centered"> | |
<!-- Lifecycle --> | |
<div class="column"> | |
<div class="content"> | |
<h2 class="title is-3">The MPCA Lifecycle</h2> | |
<ol> | |
<li><strong>Genesis (Building):</strong> The <code>Builder</code> creates a Mycelium from a data source, performing grammatical parsing to construct a graph of nodes and relationships, which is then saved as a <code>.spore</code> file.</li> | |
<li><strong>Awakening (Loading):</strong> The system loads a <code>.spore</code> file into memory, awakening the Cognitive Mycelium.</li> | |
<li><strong>Interaction (Solving):</strong> A user asks a question. The <code>Solver</code> takes the core concepts and finds a thought-path through the Mycelium to construct an answer.</li> | |
<li><strong>Evolution (Learning):</strong> Successful thought-paths are reinforced via Hebbian learning, strengthening the system's knowledge.</li> | |
<li><strong>Introspection (Dreaming):</strong> During downtime, the system can dream to form new, speculative connections, expanding its creative potential.</li> | |
</ol> | |
</div> | |
</div> | |
<!--/ Lifecycle. --> | |
</div> | |
<!-- The Path Forward --> | |
<div class="columns is-centered"> | |
<div class="column is-full-width"> | |
<h2 class="title is-3">The Path Forward: Extending MPCA to a Multi-Modal World</h2> | |
<div class="content has-text-justified"> | |
<p> | |
The current architecture is a powerful foundation for understanding language. Its true potential lies in extending this conceptual graph to understand and generate other forms of data. The core principle is that MPCA acts as a <strong>central orchestrator</strong>, connecting its abstract conceptual understanding to specialized external models for processing and generation. | |
</p> | |
</div> | |
<h3 class="title is-4">Image Understanding and Generation</h3> | |
<div class="content has-text-justified"> | |
<p><strong>Understanding:</strong> A Vision-Language Model (VLM) like CLIP analyzes an image and outputs concept tags (e.g., "a red car on a street"). The <code>Builder</code> integrates these concepts into the Mycelium, linking an <code>ImageNode</code> to existing nodes like <code>Node('car')</code> and <code>Node('red')</code>. The graph learns <em>what's in the image</em>, not the pixels themselves.</p> | |
<p><strong>Generation:</strong> The <code>Solver</code> assembles a conceptual blueprint (e.g., <code>Node('boat') → Node('blue') → Node('ocean')</code>). This blueprint is passed as a highly-structured prompt to an external image generation model (like a VAE or Diffusion model) to render the final image.</p> | |
</div> | |
<h3 class="title is-4">Audio & Video</h3> | |
<div class="content has-text-justified"> | |
<p> | |
A similar approach applies to audio and video. For audio, speech-to-text models provide text for integration, while event detection models can identify non-speech sounds ("dog barking"). For video, an analysis model tracks objects and actions over time, which the <code>Builder</code> represents as a complex, time-stamped sub-graph. | |
</p> | |
</div> | |
<h3 class="title is-4">Taking Action (Agency)</h3> | |
<div class="content has-text-justified"> | |
<p> | |
Actions are a native <code>Node</code> type in MPCA. To enable agency, these action nodes can be linked to real-world API calls or robotic functions. When the <code>Solver</code>'s thought-path traverses an <code>ActionNode</code> linked to an external function (e.g., <code>Node('turn_on_light')</code>), it triggers that function. This turns the MPCA from a passive knowledge base into an active agent that can perceive, reason about, and act upon its environment. | |
</p> | |
</div> | |
</div> | |
</div> | |
<!--/ The Path Forward --> | |
</div> | |
</section> | |
<footer class="footer"> | |
<div class="container"> | |
<div class="content has-text-centered"> | |
<a class="icon-link" href="https://github.com/bahira/myco-physarum" target="_blank" class="external-link" disabled> | |
<i class="fab fa-github"></i> | |
</a> | |
</div> | |
<div class="columns is-centered"> | |
<div class="column is-8"> | |
<div class="content"> | |
<p> | |
This website is licensed under a <a rel="license" target="_blank" | |
href="http://creativecommons.org/licenses/by-sa/4.0/">Creative | |
Commons Attribution-ShareAlike 4.0 International License</a>. | |
</p> | |
<p> | |
This page template was borrowed from the <a target="_blank" href="https://github.com/nerfies/nerfies.github.io">Nerfies</a> project website. | |
</p> | |
</div> | |
</div> | |
</div> | |
</div> | |
</footer> | |
</body> | |
</html> |