Attention, Memory, Reasoning, Continual

What this subject is for: How models attend, remember, and reason — attention, retrieval, long context, in-context learning, multi-step reasoning, continual learning.

Track status: 6 substantive concept pages. See the live generation status and the latest retrospective.

Concepts

• Attention
• In-context learning
• Long Context
• Multi-Head Attention
• Positional Encoding
• Retrieval-Augmented Generation

Arcs through this subject

No arcs yet — the retrospective proposes these once concept coverage hits ≥4 pages per track.

Key thinkers

Author pages pending.

Builds tied to this subject

MVB recipes pending — currently they live inside concept pages' Build it sections.

---

Auto-rebuilt from filesystem state by scripts/rebuild_track_indexes.py — see system architecture.