The Knowledge Tab visualizes the mindX knowledge graph, showing semantic relationships between beliefs, goals, strategic evolution, and learned patterns.
Status: ✅ DEPLOYED & OPERATIONAL Features: Interactive knowledge graph, belief system visualization, evolution tracking Backend: pgvectorscale semantic memory with vector similarity search
┌─────────────────────────────────┐
│ 💭 BELIEF: Self-Improvement │
│ Confidence: 94.7% │
│ Source: EXPERIENCE │
│ Created: 2026-01-15 │
│ Evidence: 127 memories │
└─────────────────────────────────┘
┌─────────────────────────────────┐
│ 🎯 GOAL: Optimize Performance │
│ Priority: HIGH │
│ Status: IN_PROGRESS │
│ Progress: 67% │
│ Deadline: 2026-02-01 │
└─────────────────────────────────┘
┌─────────────────────────────────┐
│ 📋 STRATEGY: Memory-Driven │
│ Type: IMPROVEMENT │
│ Phase: IMPLEMENTATION │
│ Success Rate: 89% │
└─────────────────────────────────┘
┌─────────────────────────────────┐
│ 🔄 PATTERN: High-Load Handling │
│ Occurrences: 23 │
│ Confidence: 87.3% │
│ Last Seen: 2 hours ago │
└─────────────────────────────────┘
Confidence = Base_Evidence × Source_Weight × Temporal_Decay × Reinforcement_Factor
Where:
- Base_Evidence: Number of supporting memories
- Source_Weight: Credibility of evidence source
- Temporal_Decay: Recency factor
- Reinforcement_Factor: Repeated confirmation bonus
1. Initial Formation
↓ Evidence accumulation
- Confidence Growth
↓ Pattern recognition
- Belief Refinement
↓ Contradiction resolution
- Stable Belief
↓ Periodic validation
- Belief Update/Deprecation
class KnowledgeTab extends TabComponent {
constructor(config) {
super({
id: 'knowledge',
label: 'Knowledge',
refreshInterval: 30000,
autoRefresh: true
});
}
renderKnowledgeGraph(data) {
// Initialize D3.js force-directed graph
const simulation = d3.forceSimulation(data.nodes)
.force('link', d3.forceLink(data.edges).id(d => d.id))
.force('charge', d3.forceManyBody().strength(-300))
.force('center', d3.forceCenter(width / 2, height / 2));
// Render nodes and edges
this.renderNodes(data.nodes);
this.renderEdges(data.edges);
}
}
GET /knowledge/graph
Response: {
"nodes": [
{
"id": "belief_001",
"type": "belief",
"label": "Self-Improvement Capability",
"confidence": 0.947,
"evidence_count": 127
}
],
"edges": [
{
"source": "belief_001",
"target": "goal_003",
"relationship": "supports",
"strength": 0.85
}
]
}
GET /knowledge/beliefs
Response: {
"beliefs": [
{
"belief_id": "belief_001",
"content": "System can improve through memory-driven feedback",
"confidence": 0.947,
"source": "EXPERIENCE",
"evidence": ["mem_001", "mem_002", ...]
}
]
}
GET /knowledge/goals
Response: {
"goals": [
{
"goal_id": "goal_003",
"description": "Optimize query response time",
"priority": "HIGH",
"status": "IN_PROGRESS",
"progress": 0.67
}
]
}
-- Find related beliefs by semantic similarity
SELECT b.belief_id, b.content, 1 - (e.embedding <=> query_embedding) as similarity
FROM beliefs b
JOIN belief_embeddings e ON b.belief_id = e.belief_id
WHERE 1 - (e.embedding <=> query_embedding) > 0.7
ORDER BY similarity DESC
LIMIT 10;
# Discover new patterns from memory
async def discover_patterns(memories: List[Memory]) -> List[Pattern]:
# Generate embeddings for recent memories
embeddings = [model.encode(m.content) for m in memories]
# Cluster similar memories
clusters = cluster_embeddings(embeddings, min_cluster_size=5)
# Extract patterns from clusters
patterns = []
for cluster in clusters:
pattern = extract_pattern(cluster)
if pattern.confidence > 0.7:
patterns.append(pattern)
return patterns
The Knowledge Tab provides visibility into the mindX cognitive architecture, enabling understanding and optimization of the system's belief structures and learning processes.