fix(ti): lower default lr/batch, add lr_batch sweep group

n4_baseline showed token_norm growing linearly without plateau — classic
sign of lr too high relative to parameter count. With only K×1024 params,
gradient signal per param is already high-magnitude; high lr causes
overshoot rather than convergence.

- Default lr: 1e-3 → 2e-4 (matches LoRA working regime)
- Default batch_size: 16 → 4 (more diverse gradients, helps norm saturate)
- ti_sweep_1.json: add lr_batch group (lr_low_b4, lr_mid_b8,
  lr_low_b4_prefix, lr_2e3), restructure with clearer groups,
  annotate n4_baseline as completed with findings

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

nodes/selva_ti_scheduler.py

@@ -75,9 +75,9 @@ def _get_system_info() -> dict:
 
 _PARAM_DEFAULTS = {
     "n_tokens":     4,
-    "lr":           1e-3,
+    "lr":           2e-4,
     "steps":        3000,
-    "batch_size":   16,
+    "batch_size":   4,
     "warmup_steps": 100,
     "seed":         42,
     "save_every":   1000,