fix(v8): increase micro_reward_scale 0.001→0.01 (10x stronger bootstrap signal)

At 0.001, the dense micro-reward was 1/1000th the magnitude of the
sparse exit reward (±10). Too weak to meaningfully bootstrap Q-values.
At 0.01, it's 1/100th — strong enough to provide directional feedback
while still subordinate to the exit signal.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

crates/ml/src/trainers/dqn/config.rs

@@ -1656,7 +1656,7 @@ impl DQNHyperparameters {
             exposure_aux_weight: 0.5,
             exposure_aux_warmup_epochs: 5,
             // v8: Dense micro-reward, TD(λ), PopArt, curriculum, hindsight
-            micro_reward_scale: 0.001,
+            micro_reward_scale: 0.01,
             td_lambda: 0.9,
             max_trace_length: 7,
             popart_enabled: false,  // disabled by default until wired