Update TODO.md — σ=0 achieved, all blockers resolved, document findings"

TODO.md

@@ -1,42 +1,73 @@
 # TODO (Prioritised)
 
-## Immediate (blockers)
-- ~~**Add candidate snapshot to beam logs**~~ ✅ Done — `beam_logging.py` logs candidates; `fix_and_inspect_logs.py` attaches `candidate_array` post‑hoc.
-- ~~**Ensure gate values are booleans**~~ ✅ Done — `coerce_gates()` in `fix_and_inspect_logs.py` and `postprocess_logs.py`.
-- ~~**Make tile transform nontrivial**~~ ⚠️ Partial — `ShiftedTile` exists in `transforms.py` but is **not wired into `default_atomic_factory`**. The beam only sees vanilla `tile_to_target` (idempotent), so σ never decreases. **This is the critical blocker.**
-- ~~**Implement robust fill_enclosed**~~ ✅ Done — BFS implementation in `solver_core.py`.
+## ✅ σ=0 ACHIEVED — Solver works!
 
-### ➡️ Next: Fix σ=98 flatline (residue never decreases)
-The beam composes `tile_to_target ∘ tile_to_target` which is idempotent — tiling an already-tiled grid returns the same grid. The atomic library needs transforms that can actually express the input→target mapping.
+The solver now finds `Id∘KroneckerSelfSimilar` at depth 1 and achieves **σ=0** on all 6 pairs of ARC task 007bbfb7.
 
-**Plan:**
-1. **Wire `ShiftedTile` into `default_atomic_factory`** so the beam can explore shifted tilings.
-2. **Add region‑aware placement transforms** — the example1 target is a 3×3 arrangement of sub‑blocks where each 3×3 quadrant is a conditional variant of the input (some reflected, some identity, some zero‑filled). The library needs transforms that place sub‑blocks at specific offsets.
-3. **Add conditional fill / mask‑based composition** — transforms that zero‑out or overwrite specific quadrants based on input symmetry.
-4. **Re‑run sweep** with enriched library and verify σ decreases below 98.
+### What was wrong
+1. **Wrong target** — the repo's example1 target had 4 incorrect cells (row 7). Discovered by cross-referencing against the real ARC dataset (`data/training/007bbfb7.json`). Fixed.
+2. **Limited transform library** — the beam only had vanilla tile, fill_enclosed, rotate90, reflect_h. None could express the Kronecker self-similar pattern. Fixed: added 19 new transforms.
+3. **Beam only tried resized input** — shape-changing transforms (Kronecker: 3×3→9×9) need the original input, not the tiled 9×9 intermediate. The beam now uses a dual-strategy: each transform is tried on both the resized field AND the original input.
+
+### The actual transformation (ARC task 007bbfb7)
+`output = np.kron((input != 0).astype(int), input)` — a Kronecker product where the input's own nonzero mask determines the meta-layout for placing copies of itself.
+
+---
+
+## Immediate (blockers) — ALL RESOLVED
+- ~~**Add candidate snapshot to beam logs**~~ ✅ Done.
+- ~~**Ensure gate values are booleans**~~ ✅ Done.
+- ~~**Make tile transform nontrivial**~~ ✅ Done — `ShiftedTile` wired in + 19 new transforms including `KroneckerSelfSimilar`.
+- ~~**Implement robust fill_enclosed**~~ ✅ Done — BFS in `solver_core.py`.
+- ~~**Fix σ=98 flatline**~~ ✅ Done — σ=0 on all 6 pairs.
 
 ## Short term
 - ~~Add CLI entrypoint with `--use_wandb` flag~~ ✅ Done — `scripts/entrypoint.py`.
-- Add unit tests for `tile_transform`, `fill_enclosed`, and transform `.apply()` semantics.  
-  - `tests.py` has `run_atomic_effects` and `transform_effect_test` but no proper pytest suite yet.
+- ~~Add unit tests for transforms~~ ✅ Done — `tests/test_transforms.py` (40 tests, all pass).
 - Add small visualization notebook for `phi_best`, diff maps, and Layer‑1 masks.
 
 ## Medium term
-- ~~Improve Layer‑1 mask generation~~ ✅ Done — percentile + min_abs + dilation in `layer_minus_one.py`.
+- ~~Improve Layer‑1 mask generation~~ ✅ Done.
 - Add a toggle to include/exclude `candidate_array` in logs to control log size.
-- ~~Create a reproducible benchmark harness for parameter sweeps and CSV aggregation~~ ✅ Done — `experiment_driver.sweep()` + `results.csv`. Needs more tasks.
+- ~~Create a reproducible benchmark harness~~ ✅ Done — `experiment_driver.sweep()` + `results.csv`.
+- **Expand to more ARC tasks** — test on other 3×3→9×9 tasks and different task families.
+- **Benchmark the enriched library** — run sweep across multiple ARC tasks, measure solve rate.
 
 ## Long term
-- ~~Integrate a safe external W&B uploader~~ ✅ Done — `wandb_runner.py` runs after experiments finish, decoupled from core.
-- Add more ARC tasks and automated evaluation harness.
+- ~~Integrate a safe external W&B uploader~~ ✅ Done.
+- **Build task loader for full ARC dataset** — load any task from `fchollet/ARC-AGI` by ID.
+- **Add more transform families** — connected components, object extraction, voronoi fill (see Icecuber DSL: arxiv:2402.03507).
+- **Automated evaluation harness** — run solver on all 400 ARC training tasks, report solve rate.
 - Document reproducibility steps and expected outputs for each example task.
 
 ## Code hygiene (completed)
-- ~~Duplicate `Transform` class in `transforms.py`~~ ✅ Fixed — now imports from `solver_core`.
-- ~~Duplicate imports/paste blocks in `solver_core.py`~~ ✅ Fixed — single clean definition.
-- ~~Lambda closure bug in `default_atomic_factory`~~ ✅ Fixed — captures `target_shape` by value.
-- ~~`wandb_runner.py` `int(generate_id(), 36)` crash~~ ✅ Fixed — uses string ID directly.
-- ~~`minimal_runner.py` TARGET was all‑zeros~~ ✅ Fixed — uses real example1 target.
-- ~~README.md referenced non‑existent paths~~ ✅ Fixed — corrected to actual repo structure.
-- ~~Committed `.pyc` files~~ ✅ Fixed — deleted + `.gitignore` added.
-- ~~`itt_solver/README.md.md` double extension~~ ✅ Fixed — renamed to `README.md`.
+- ~~Duplicate `Transform` class in `transforms.py`~~ ✅ Fixed.
+- ~~Duplicate imports/paste blocks in `solver_core.py`~~ ✅ Fixed.
+- ~~Lambda closure bug in `default_atomic_factory`~~ ✅ Fixed.
+- ~~`wandb_runner.py` `int(generate_id(), 36)` crash~~ ✅ Fixed.
+- ~~`minimal_runner.py` TARGET was all‑zeros~~ ✅ Fixed.
+- ~~README.md referenced non‑existent paths~~ ✅ Fixed.
+- ~~Committed `.pyc` files~~ ✅ Fixed.
+- ~~`itt_solver/README.md.md` double extension~~ ✅ Fixed.
+- ~~Wrong target in example1 (4 cells off)~~ ✅ Fixed — corrected in entrypoint.py, experiments_analysis.py, fix_and_inspect_logs.py, minimal_runner.py.
+- ~~Beam search only applied transforms to resized field~~ ✅ Fixed — dual-strategy (resized + original).
+
+## New transforms added (19 total)
+| Transform | Description |
+|---|---|
+| `KroneckerSelfSimilar` | `kron((I≠0), I)` — self-similar meta-layout |
+| `KroneckerSelfSimilarInv` | `kron(I, (I≠0))` — mirror variant |
+| `MirrorTileH` | `[abc\|cba]` horizontal mirror |
+| `MirrorTileV` | Vertical mirror stack |
+| `MirrorTile4Way` | Full kaleidoscope (D4) |
+| `Upscale(2)` / `Upscale(3)` | Pixel-repeat zoom |
+| `Downscale(2)` | Subsample (inverse of upscale) |
+| `StackH(3)` / `StackV(3)` | Tile horizontally/vertically |
+| `RetainColor(c)` | Keep only color c |
+| `RemoveColor(c)` | Zero out color c |
+| `InvertColors` | Swap black ↔ top color |
+| `GravityDown` / `GravityUp` | Pixels fall/rise in columns |
+| `OverlayTransparent(bg)` | Transparent overlay on background |
+| `CropToContent` | Crop to non-zero bounding box |
+| `Transpose` | Matrix transpose |
+| `ShiftedTile` | Tile with roll offset |