Spaces:
Running
Running
File size: 3,608 Bytes
3d46076 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 | from dataclasses import dataclass, field
import numpy as np
from typing import Dict, Any, List, Optional
@dataclass
class HomeostaticDrives:
energy: float = 1.0 # Depletes over time, replenished by sleep/reward
curiosity: float = 0.8 # Increases with novel inputs, decreases with habituation
social: float = 0.5 # Drives communication / speech
integrity: float = 1.0 # Reduced by error/damage, recovered by rest
def step(self, activity_level: float, novelty: float):
# Metabolism consumes energy
self.energy = max(0.0, min(1.0, self.energy - 0.001 * (1.0 + activity_level)))
# Curiosity increases with novelty, decays with stagnation
self.curiosity = max(0.0, min(1.0, self.curiosity * 0.99 + 0.1 * novelty))
# Social drive gently drifts up to encourage interaction
self.social = max(0.0, min(1.0, self.social + 0.0005))
# Integrity recovers if energy is adequate
if self.energy > 0.3:
self.integrity = min(1.0, self.integrity + 0.001)
@dataclass
class BrainState:
num_neurons: int
membrane_potentials: np.ndarray # [N] float32
spikes: np.ndarray # [N] float32 (0.0 or 1.0)
refractory_steps: np.ndarray # [N] int32
activations: np.ndarray # [N] float32 (filtered rate / display)
attention: np.ndarray # [N] float32 attention focus
prediction_error: float = 0.0
predicted_reward: float = 0.0
current_reward: float = 0.0
drives: HomeostaticDrives = field(default_factory=HomeostaticDrives)
active_goal: str = "explore"
goal_embedding: np.ndarray = field(default_factory=lambda: np.zeros(16, dtype=np.float32))
tool_associations: Dict[str, float] = field(default_factory=dict)
active_memory_refs: List[str] = field(default_factory=list)
step_count: int = 0
total_spikes: int = 0
@classmethod
def create_initial(cls, num_neurons: int, seed: int = 42) -> 'BrainState':
rng = np.random.RandomState(seed)
pot = rng.uniform(0.0, 0.2, num_neurons).astype(np.float32)
spikes = np.zeros(num_neurons, dtype=np.float32)
ref = np.zeros(num_neurons, dtype=np.int32)
act = np.zeros(num_neurons, dtype=np.float32)
att = np.ones(num_neurons, dtype=np.float32) / num_neurons
return cls(
num_neurons=num_neurons,
membrane_potentials=pot,
spikes=spikes,
refractory_steps=ref,
activations=act,
attention=att,
goal_embedding=np.zeros(16, dtype=np.float32)
)
def to_dict(self) -> Dict[str, Any]:
return {
"step_count": self.step_count,
"total_spikes": self.total_spikes,
"prediction_error": float(self.prediction_error),
"predicted_reward": float(self.predicted_reward),
"current_reward": float(self.current_reward),
"active_goal": self.active_goal,
"drives": {
"energy": float(self.drives.energy),
"curiosity": float(self.drives.curiosity),
"social": float(self.drives.social),
"integrity": float(self.drives.integrity)
},
"tool_associations": dict(self.tool_associations),
"active_memory_refs": list(self.active_memory_refs),
"mean_activation": float(np.mean(self.activations)),
"max_activation": float(np.max(self.activations)),
"active_spikes_count": int(np.sum(self.spikes > 0.5))
}
|