Giorgio Gilestro
840b6b00b3
Re-scopes Layer 2 into a cheaper, architecture-general neural collapse proof
before the LLM rung. Realises the same Wright–Fisher abstractions in real trained
generative models on a fully-synthetic sandbox with an exact oracle, reusing
knowledge.metrics/truth/seeding and the output contract so neural curves overlay
the Layer-1 analytic curves.
- src/neural/: synthetic token-grammar sandbox (lossless identity + stochastic
style), ExactOracle, HistogramModel bridge, generation loop, experiment runner
- HARD GATE passed: histogram lineage reproduces Layer 1 exactly (neutral decay,
exact H_eq, tracks run_lineage) — tests/test_neural_validation.py
- torch models: autoregressive RNN + MLP (VAE implemented, not yet fidelity-
passing); determinism seeding derived from the SeedSequence stream
- N0 bridge (neural g*=0.047 ≈ Layer-1 0.048), N1 collapse-in-weights, N2 phase
boundary, N5 architecture-generality (collapse + grounding-rescue in histogram
+ RNN + MLP). Manifests/configs committed; parquet gitignored, hashes tracked
- additive backward-compatible save_artifacts extension; Makefile neural targets
Finding: neural smoothing partially resists H-collapse, so forward-KL and tail
survival are the sharp neural collapse metrics (H is smooth, per Layer 1).
92 tests green. Remaining (tasks/todo.md): N4 merge, N2 refine, N3/N6, VAE
fidelity, MNIST tier, figures. LLM/LoRA rung and C3 deferred.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
13 lines
755 B
Python
13 lines
755 B
Python
"""Layer 1.5 — the architecture-general neural existence proof.
|
|
|
|
Realises the Layer-1 (``knowledge``) Wright-Fisher abstractions in *real trained
|
|
generative models* on a fully-synthetic sandbox whose ground-truth ``p*`` is known
|
|
exactly. A model's knowledge is measured as its output distribution over ``K`` discrete
|
|
*modes* (via an oracle), so the same metrics (``knowledge.metrics``), the same closed
|
|
forms, and the same experiments carry over — a neural collapse curve can be overlaid on
|
|
a Layer-1 analytic curve.
|
|
|
|
The package is staged by cost: the histogram model (pure NumPy) reduces this layer
|
|
*exactly* to Layer 1 and is the validation bridge; the RNN/VAE/MLP models (torch, added
|
|
from Stage C) show that collapse is architecture-general.
|
|
"""
|