Skip to main content

Viveka Layer

The intelligent decision-making layer that determines what to remember and what to forget.

Overview​

Viveka (Sanskrit: विवेक, "discriminative wisdom") is the layer that applies intelligent filtering and importance scoring to memories.

Importance Scoring​

Each memory receives an importance score (0.0 to 1.0) based on:

  • Recency: More recent memories score higher
  • Frequency: Often-accessed memories score higher
  • Semantic significance: Content-based importance
  • User signals: Explicit importance markers

Decay Mechanism​

from vidurai import Vidurai

# Enable decay (default)
memory = Vidurai(enable_decay=True)

# Disable decay for critical applications
memory = Vidurai(enable_decay=False)

Decay formula:

importance(t) = importance(0) * exp(-decay_rate * time_elapsed)

Filtering Strategies​

Threshold-based​

# Recall only high-importance memories
results = memory.recall(min_importance=0.7)

Count-based​

# Get top N most important
results = memory.recall(limit=10)

Time-based​

# Memories from last hour
from datetime import datetime, timedelta
cutoff = datetime.now() - timedelta(hours=1)
results = memory.recall(since=cutoff)

See Memory Management for advanced techniques.