Sovereign Nexus: Inject GOFAI Logic & AdaptiveCalibrator

app.js

@@ -76,6 +76,473 @@ const orbitState = {
 let flyY = 2;
 
 // ═══ INIT ═══
+
+// ═══ SOVEREIGN SYMBOLIC ENGINE (GOFAI) ═══
+class SymbolicEngine {
+  constructor() {
+    this.facts = new Map();
+    this.factIndices = new Map();
+    this.factMask = 0n;
+    this.rules = [];
+    this.reasoningLog = [];
+  }
+
+  addFact(key, value) {
+    this.facts.set(key, value);
+    if (!this.factIndices.has(key)) {
+      this.factIndices.set(key, BigInt(this.factIndices.size));
+    }
+    const bitIndex = this.factIndices.get(key);
+    if (value) {
+      this.factMask |= (1n << bitIndex);
+    } else {
+      this.factMask &= ~(1n << bitIndex);
+    }
+  }
+
+  addRule(condition, action, description) {
+    this.rules.push({ condition, action, description });
+  }
+
+  reason() {
+    this.rules.forEach(rule => {
+      if (rule.condition(this.facts)) {
+        const result = rule.action(this.facts);
+        if (result) {
+          this.logReasoning(rule.description, result);
+        }
+      }
+    });
+  }
+
+  logReasoning(ruleDesc, outcome) {
+    const entry = { timestamp: Date.now(), rule: ruleDesc, outcome: outcome };
+    this.reasoningLog.unshift(entry);
+    if (this.reasoningLog.length > 5) this.reasoningLog.pop();
+    
+    const container = document.getElementById('symbolic-log-content');
+    if (container) {
+      const logDiv = document.createElement('div');
+      logDiv.className = 'symbolic-log-entry';
+      logDiv.innerHTML = `<span class="symbolic-rule">[RULE] ${ruleDesc}</span><span class="symbolic-outcome">→ ${outcome}</span>`;
+      container.prepend(logDiv);
+      if (container.children.length > 5) container.lastElementChild.remove();
+    }
+  }
+}
+
+class AgentFSM {
+  constructor(agentId, initialState) {
+    this.agentId = agentId;
+    this.state = initialState;
+    this.transitions = {};
+  }
+
+  addTransition(fromState, toState, condition) {
+    if (!this.transitions[fromState]) this.transitions[fromState] = [];
+    this.transitions[fromState].push({ toState, condition });
+  }
+
+  update(facts) {
+    const possibleTransitions = this.transitions[this.state] || [];
+    for (const transition of possibleTransitions) {
+      if (transition.condition(facts)) {
+        console.log(`[FSM] Agent ${this.agentId} transitioning: ${this.state} -> ${transition.toState}`);
+        this.state = transition.toState;
+        return true;
+      }
+    }
+    return false;
+  }
+}
+
+class KnowledgeGraph {
+  constructor() {
+    this.nodes = new Map();
+    this.edges = [];
+  }
+
+  addNode(id, type, metadata = {}) {
+    this.nodes.set(id, { id, type, ...metadata });
+  }
+
+  addEdge(from, to, relation) {
+    this.edges.push({ from, to, relation });
+  }
+
+  query(from, relation) {
+    return this.edges
+      .filter(e => e.from === from && e.relation === relation)
+      .map(e => this.nodes.get(e.to));
+  }
+}
+
+class Blackboard {
+  constructor() {
+    this.data = {};
+    this.subscribers = [];
+  }
+
+  write(key, value, source) {
+    const oldValue = this.data[key];
+    this.data[key] = value;
+    this.notify(key, value, oldValue, source);
+  }
+
+  read(key) { return this.data[key]; }
+
+  subscribe(callback) { this.subscribers.push(callback); }
+
+  notify(key, value, oldValue, source) {
+    this.subscribers.forEach(sub => sub(key, value, oldValue, source));
+    const container = document.getElementById('blackboard-log-content');
+    if (container) {
+      const entry = document.createElement('div');
+      entry.className = 'blackboard-entry';
+      entry.innerHTML = `<span class="bb-source">[${source}]</span> <span class="bb-key">${key}</span>: <span class="bb-value">${JSON.stringify(value)}</span>`;
+      container.prepend(entry);
+      if (container.children.length > 8) container.lastElementChild.remove();
+    }
+  }
+}
+
+class SymbolicPlanner {
+  constructor() {
+    this.actions = [];
+    this.currentPlan = [];
+  }
+
+  addAction(name, preconditions, effects) {
+    this.actions.push({ name, preconditions, effects });
+  }
+
+  heuristic(state, goal) {
+    let h = 0;
+    for (let key in goal) {
+      if (state[key] !== goal[key]) {
+        h += Math.abs((state[key] || 0) - (goal[key] || 0));
+      }
+    }
+    return h;
+  }
+
+  findPlan(initialState, goalState) {
+    let openSet = [{ state: initialState, plan: [], g: 0, h: this.heuristic(initialState, goalState) }];
+    let visited = new Map();
+    visited.set(JSON.stringify(initialState), 0);
+
+    while (openSet.length > 0) {
+      openSet.sort((a, b) => (a.g + a.h) - (b.g + b.h));
+      let { state, plan, g } = openSet.shift();
+
+      if (this.isGoalReached(state, goalState)) return plan;
+
+      for (let action of this.actions) {
+        if (this.arePreconditionsMet(state, action.preconditions)) {
+          let nextState = { ...state, ...action.effects };
+          let stateStr = JSON.stringify(nextState);
+          let nextG = g + 1;
+
+          if (!visited.has(stateStr) || nextG < visited.get(stateStr)) {
+            visited.set(stateStr, nextG);
+            openSet.push({
+              state: nextState,
+              plan: [...plan, action.name],
+              g: nextG,
+              h: this.heuristic(nextState, goalState)
+            });
+          }
+        }
+      }
+    }
+    return null;
+  }
+
+  isGoalReached(state, goal) {
+    for (let key in goal) {
+      if (state[key] !== goal[key]) return false;
+    }
+    return true;
+  }
+
+  arePreconditionsMet(state, preconditions) {
+    for (let key in preconditions) {
+      if (state[key] < preconditions[key]) return false;
+    }
+    return true;
+  }
+
+  logPlan(plan) {
+    this.currentPlan = plan;
+    const container = document.getElementById('planner-log-content');
+    if (container) {
+      container.innerHTML = '';
+      if (!plan || plan.length === 0) {
+        container.innerHTML = '<div class="planner-empty">NO ACTIVE PLAN</div>';
+        return;
+      }
+      plan.forEach((step, i) => {
+        const div = document.createElement('div');
+        div.className = 'planner-step';
+        div.innerHTML = `<span class="step-num">${i+1}.</span> ${step}`;
+        container.appendChild(div);
+      });
+    }
+  }
+}
+
+class HTNPlanner {
+  constructor() {
+    this.methods = {};
+    this.primitiveTasks = {};
+  }
+
+  addMethod(taskName, preconditions, subtasks) {
+    if (!this.methods[taskName]) this.methods[taskName] = [];
+    this.methods[taskName].push({ preconditions, subtasks });
+  }
+
+  addPrimitiveTask(taskName, preconditions, effects) {
+    this.primitiveTasks[taskName] = { preconditions, effects };
+  }
+
+  findPlan(initialState, tasks) {
+    return this.decompose(initialState, tasks, []);
+  }
+
+  decompose(state, tasks, plan) {
+    if (tasks.length === 0) return plan;
+    const [task, ...remainingTasks] = tasks;
+    if (this.primitiveTasks[task]) {
+      const { preconditions, effects } = this.primitiveTasks[task];
+      if (this.arePreconditionsMet(state, preconditions)) {
+        const nextState = { ...state, ...effects };
+        return this.decompose(nextState, remainingTasks, [...plan, task]);
+      }
+      return null;
+    }
+    const methods = this.methods[task] || [];
+    for (const method of methods) {
+      if (this.arePreconditionsMet(state, method.preconditions)) {
+        const result = this.decompose(state, [...method.subtasks, ...remainingTasks], plan);
+        if (result) return result;
+      }
+    }
+    return null;
+  }
+
+  arePreconditionsMet(state, preconditions) {
+    for (const key in preconditions) {
+      if (state[key] < (preconditions[key] || 0)) return false;
+    }
+    return true;
+  }
+}
+
+class CaseBasedReasoner {
+  constructor() {
+    this.caseLibrary = [];
+  }
+
+  addCase(situation, action, outcome) {
+    this.caseLibrary.push({ situation, action, outcome, timestamp: Date.now() });
+  }
+
+  findSimilarCase(currentSituation) {
+    let bestMatch = null;
+    let maxSimilarity = -1;
+    this.caseLibrary.forEach(c => {
+      let similarity = this.calculateSimilarity(currentSituation, c.situation);
+      if (similarity > maxSimilarity) {
+        maxSimilarity = similarity;
+        bestMatch = c;
+      }
+    });
+    return maxSimilarity > 0.7 ? bestMatch : null;
+  }
+
+  calculateSimilarity(s1, s2) {
+    let score = 0, total = 0;
+    for (let key in s1) {
+      if (s2[key] !== undefined) {
+        score += 1 - Math.abs(s1[key] - s2[key]);
+        total += 1;
+      }
+    }
+    return total > 0 ? score / total : 0;
+  }
+
+  logCase(c) {
+    const container = document.getElementById('cbr-log-content');
+    if (container) {
+      const div = document.createElement('div');
+      div.className = 'cbr-entry';
+      div.innerHTML = `
+        <div class="cbr-match">SIMILAR CASE FOUND (${(this.calculateSimilarity(symbolicEngine.facts, c.situation) * 100).toFixed(0)}%)</div>
+        <div class="cbr-action">SUGGESTED: ${c.action}</div>
+        <div class="cbr-outcome">PREVIOUS OUTCOME: ${c.outcome}</div>
+      `;
+      container.prepend(div);
+      if (container.children.length > 3) container.lastElementChild.remove();
+    }
+  }
+}
+
+class NeuroSymbolicBridge {
+  constructor(symbolicEngine, blackboard) {
+    this.engine = symbolicEngine;
+    this.blackboard = blackboard;
+    this.perceptionLog = [];
+  }
+
+  perceive(rawState) {
+    const concepts = [];
+    if (rawState.stability < 0.4 && rawState.energy > 60) concepts.push('UNSTABLE_OSCILLATION');
+    if (rawState.energy < 30 && rawState.activePortals > 2) concepts.push('CRITICAL_DRAIN_PATTERN');
+    concepts.forEach(concept => {
+      this.engine.addFact(concept, true);
+      this.logPerception(concept);
+    });
+    return concepts;
+  }
+
+  logPerception(concept) {
+    const container = document.getElementById('neuro-bridge-log-content');
+    if (container) {
+      const div = document.createElement('div');
+      div.className = 'neuro-bridge-entry';
+      div.innerHTML = `<span class="neuro-icon">🧠</span> <span class="neuro-concept">${concept}</span>`;
+      container.prepend(div);
+      if (container.children.length > 5) container.lastElementChild.remove();
+    }
+  }
+}
+
+class MetaReasoningLayer {
+  constructor(planner, blackboard) {
+    this.planner = planner;
+    this.blackboard = blackboard;
+    this.reasoningCache = new Map();
+    this.performanceMetrics = { totalReasoningTime: 0, calls: 0 };
+  }
+
+  getCachedPlan(stateKey) {
+    const cached = this.reasoningCache.get(stateKey);
+    if (cached && (Date.now() - cached.timestamp < 10000)) return cached.plan;
+    return null;
+  }
+
+  cachePlan(stateKey, plan) {
+    this.reasoningCache.set(stateKey, { plan, timestamp: Date.now() });
+  }
+
+  reflect() {
+    const avgTime = this.performanceMetrics.totalReasoningTime / (this.performanceMetrics.calls || 1);
+    const container = document.getElementById('meta-log-content');
+    if (container) {
+      container.innerHTML = `
+        <div class="meta-stat">CACHE SIZE: ${this.reasoningCache.size}</div>
+        <div class="meta-stat">AVG LATENCY: ${avgTime.toFixed(2)}ms</div>
+        <div class="meta-stat">STATUS: ${avgTime > 50 ? 'OPTIMIZING' : 'NOMINAL'}</div>
+      `;
+    }
+  }
+
+  track(startTime) {
+    const duration = performance.now() - startTime;
+    this.performanceMetrics.totalReasoningTime += duration;
+    this.performanceMetrics.calls++;
+  }
+}
+
+// ═══ ADAPTIVE CALIBRATOR (LOCAL EFFICIENCY) ═══
+class AdaptiveCalibrator {
+  constructor(modelId, initialParams) {
+    this.model = modelId;
+    this.weights = {
+      'input_tokens': 0.0,
+      'complexity_score': 0.0,
+      'task_type_indicator': 0.0,
+      'bias': initialParams.base_rate || 0.0
+    };
+    this.learningRate = 0.01;
+    this.history = [];
+  }
+
+  predict(features) {
+    let prediction = this.weights['bias'];
+    for (let feature in features) {
+      if (this.weights[feature] !== undefined) {
+        prediction += this.weights[feature] * features[feature];
+      }
+    }
+    return Math.max(0, prediction);
+  }
+
+  update(features, actualCost) {
+    const predicted = this.predict(features);
+    const error = actualCost - predicted;
+    for (let feature in features) {
+      if (this.weights[feature] !== undefined) {
+        this.weights[feature] += this.learningRate * error * features[feature];
+      }
+    }
+    this.history.push({ predicted, actual: actualCost, timestamp: Date.now() });
+    
+    const container = document.getElementById('calibrator-log-content');
+    if (container) {
+      const div = document.createElement('div');
+      div.className = 'calibrator-entry';
+      div.innerHTML = `<span class="cal-label">CALIBRATED:</span> <span class="cal-val">${predicted.toFixed(4)}</span> <span class="cal-err">ERR: ${error.toFixed(4)}</span>`;
+      container.prepend(div);
+      if (container.children.length > 5) container.lastElementChild.remove();
+    }
+  }
+}
+
+let metaLayer, neuroBridge, cbr, symbolicPlanner, knowledgeGraph, blackboard, symbolicEngine, calibrator;
+let agentFSMs = {};
+
+function setupGOFAI() {
+  knowledgeGraph = new KnowledgeGraph();
+  blackboard = new Blackboard();
+  symbolicEngine = new SymbolicEngine();
+  symbolicPlanner = new SymbolicPlanner();
+  cbr = new CaseBasedReasoner();
+  neuroBridge = new NeuroSymbolicBridge(symbolicEngine, blackboard);
+  metaLayer = new MetaReasoningLayer(symbolicPlanner, blackboard);
+  calibrator = new AdaptiveCalibrator('nexus-v1', { base_rate: 0.05 });
+
+  // Setup initial facts
+  symbolicEngine.addFact('energy', 100);
+  symbolicEngine.addFact('stability', 1.0);
+  
+  // Setup FSM
+  agentFSMs['timmy'] = new AgentFSM('timmy', 'IDLE');
+  agentFSMs['timmy'].addTransition('IDLE', 'ANALYZING', (facts) => facts.get('activePortals') > 0);
+  
+  // Setup Planner
+  symbolicPlanner.addAction('Stabilize Matrix', { energy: 50 }, { stability: 1.0 });
+}
+
+function updateGOFAI(delta, elapsed) {
+  const startTime = performance.now();
+  
+  // Simulate perception
+  neuroBridge.perceive({ stability: 0.3, energy: 80, activePortals: 1 });
+  
+  // Run reasoning
+  if (Math.floor(elapsed * 2) > Math.floor((elapsed - delta) * 2)) {
+    symbolicEngine.reason();
+    metaLayer.reflect();
+    
+    // Simulate calibration update
+    calibrator.update({ input_tokens: 100, complexity_score: 0.5 }, 0.06);
+  }
+  
+  metaLayer.track(startTime);
+}
+
 async function init() {
   clock = new THREE.Clock();
   playerPos = new THREE.Vector3(0, 2, 12);
@@ -95,6 +562,7 @@ async function init() {
   scene = new THREE.Scene();
   scene.fog = new THREE.FogExp2(0x050510, 0.012);
 
+  setupGOFAI();
   camera = new THREE.PerspectiveCamera(65, window.innerWidth / window.innerHeight, 0.1, 1000);
   camera.position.copy(playerPos);