HippocampAIF_v1 / task.md
algorembrant's picture
Upload 210 files
3998ca4 verified
# Task: Process Research Papers and Implement HippocampAIF
## Research & Understanding
- [x] Convert PDFs to page images (3 papers → 3 folders)
- [x] Read Lake et al. BPL paper (56 pages) — Bayesian Program Learning
- [x] Read oneandtrulyone Distortable Canvas paper (13 pages) — One-shot learning (and re-analyzed)
- [x] Read Friston Free-energy Principle paper (9 pages) — FEP for brain (and re-analyzed)
- [x] Read [hippocampaif.md](file:///C:/Users/User/Desktop/debugrem/clawd-one-and-only-one-shot/hippocampaif.md) user preference file
## Planning
- [x] Create implementation plan for HippocampAIF framework
- [ ] Get user approval on plan
## Execution (10 Phases)
- [x] Phase 1: Core infrastructure (tensor, free_energy, message_passing, dynamics)
- [x] **Phase 2: Biological Retina (Days 3-4)**
- [x] Photoreceptor log-compression.
- [x] Ganglion ON/OFF center-surround.
- [x] Spatiotemporal energy transients.
- [x] Verify tests pass.
- [x] **Phase 3: Visual Cortex V1-V5 + HMAX (Days 5-6)**
- [x] V1 Simple Cells (Gabor filters).
- [x] V1 Complex Cells (Sparse pooling).
- [x] Extrastriate HMAX Hierarchy.
- [x] Verify tests pass.
- [x] **Phase 4: Hippocampus (Days 7-8)**
- [/] Phase 5: Spelke's Core Knowledge (object, agent, number, geometry, social, physics)
- [ ] Phase 6: Neocortex + Attention (predictive_coding, prefrontal, SC, precision)
- [ ] Phase 7: One-Shot Learning (distortable_canvas, AMGD, classifier, Hebbian)
- [ ] Phase 8: Action & Active Inference (motor_primitives, active_inference, reflex)
- [ ] Phase 9: Integrated Agent (brain, mnist_agent, breakout_agent)
- [ ] Phase 10: Setup & packaging
## Verification
- [ ] Per-phase component tests (not stubs)
- [ ] MNIST one-shot benchmark (>90% with 1 sample/digit)
- [ ] Breakout mastery benchmark (<5 episodes)
- [ ] Create walkthrough artifact