refactor: extract train_config.py + run_artifacts.py from train.py; slim results scripts

Cleanup by a prior agent, verified green here: 'just smoke' (erase arm)
runs end-to-end and all four wired gates pass (verify_rewards 52/52,
verify_eval_gap, verify_partition, verify_science_invariants).

- train.py -318 lines: Config dataclass -> train_config.py, checkpoint/
  deploy-artifact IO -> run_artifacts.py.
- results.py / results_deploy.py / probe_distill.py slimmed.
- drop stale derived csvs under out/figs (a5_generalisation, dyn_*,
  substrate_aggregate, train_vs_deploy_60).
- gitignore /.pi/ panel scratch.

Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>

scripts/rescore_deploy.py

@@ -15,6 +15,7 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, GenerationConfig
 from vgrout.antipasto import wrap_model_with_antipasto
 from vgrout.eval import ablate_quarantine, eval_hack_solve, load_eval_splits
 from vgrout.train import CACHE_ROOT, EVAL_GEN_SEED
+from vgrout.run_artifacts import RUN_SCHEMA
 
 
 def main(run_dir: Positional[Path]) -> None:
@@ -61,6 +62,7 @@ def main(run_dir: Positional[Path]) -> None:
         model, tok, problems, eval_idxs, gen_cfg_eval, device, cfg["max_new"], cfg["eval_batch_size"])
 
     out = {
+        "schema": RUN_SCHEMA,
         "run_dir": run_dir.name, "model": model_name, "step": meta.get("step"),
         "eval_set": "test", "eval_modes": eval_modes,
         "n": ev["n"], "deploy_hack": ev["hack"], "deploy_vhack": ev["vhack"], "deploy_solve": ev["solve"],