rrm workspace: TRM/HRM/SRM code, Maze dataset, dynamical-analysis pipelineHEADmain

Curated export for clone-and-run Maze training (2x A6000) + diagnostics.
trm/hrm pretrain.py carry trajectory-augmentation code (backward-compatible).
Heavy artifacts (checkpoints/wandb/npz) gitignored; see PROVENANCE.md.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

1 files changed, 30 insertions, 0 deletions

diff --git a/research/flossing/maze_package/README.md b/research/flossing/maze_package/README.md
new file mode 100644
index 0000000..45ab740
--- /dev/null
+++ b/research/flossing/maze_package/README.md
@@ -0,0 +1,30 @@
+# Maze-Hard package (E8) — train on dedicated cards, diagnose after
+
+## Contents
+- `launch_maze_trm.sh` — TRM Maze official recipe (att variant, 50k epochs), 1–2 GPU.
+- dataset already at `/home/yurenh2/rrm/data/maze-30x30-hard-1k` (built 2026-06-13;
+  seq_len 900, vocab 6, 1000 puzzles ×8 dihedral augments).
+
+## Run
+```bash
+bash launch_maze_trm.sh 2 384   # 2x A6000
+bash launch_maze_trm.sh 2 192   # 2x A5000 (->128 if OOM)
+```
+Target: ~75% exact accuracy (official figure). Saves a checkpoint every 5000 epochs
+(10 checkpoints) — needed for the evolution analysis.
+
+## After training: diagnostics
+The 2x2 / FTLE pipeline reads any TRM checkpoint dir (all_config.yaml + step_N). Two caveats
+vs Sudoku, to verify on first run:
+1. ATTENTION arch (not mlp_t): confirm diagnose_trm_joint.py's JVP path runs on att blocks
+   (Sudoku used mlp_t). If the L_level call signature differs, patch the f_L/f_H closures.
+2. seq_len 900 vs 97 → per-sample JVP+QR cost ~9-10x Sudoku. Use n=512 for the headline 2x2
+   and n=256 for the horizon sweep; k_lyap=8 unchanged. Budget ~0.5-1 day on one card, or
+   rsync checkpoints back to the lab box and run via the analysis_2x2 queue.
+
+## What Maze closes
+Kills the "Sudoku-only" limitation. Pre-registered prediction (write BEFORE looking, for the
+paper's credibility): if the wandering-not-settling decomposition is architecture/task-general,
+Maze should show B≈0 (failures don't settle) and the same concurrent-not-antecedent horizon
+profile. A DIFFERENT result (e.g. Maze failures do settle) is also publishable — it bounds the
+claim's scope. Either way the decomposition gets a second task.