diff --git a/crates/ml-alpha/cuda/rl_reward_scale_controller.cu b/crates/ml-alpha/cuda/rl_reward_scale_controller.cu
index fcb98773f..58e586e56 100644
--- a/crates/ml-alpha/cuda/rl_reward_scale_controller.cu
+++ b/crates/ml-alpha/cuda/rl_reward_scale_controller.cu
@@ -130,6 +130,15 @@ extern "C" __global__ void rl_reward_scale_controller(
     const float a   = fmaxf(alpha_step, WIENER_ALPHA_FLOOR);
     float       out = (1.0f - a) * prev + a * target;
 
+    // Per-step movement clamp: scale moves at most ±2% from previous.
+    // Damps the 10,000× oscillation from sparse trade PnL spikes to
+    // a smooth ramp. At 2%/step, doubling takes ~35 steps, halving
+    // takes ~35 steps — fast enough to track regime changes, slow
+    // enough to give Q stable targets across the replay window.
+    const float max_move = prev * 1.02f;
+    const float min_move = prev * 0.98f;
+    out = fmaxf(min_move, fminf(out, max_move));
+
     out = fmaxf(scale_min, fminf(out, REWARD_SCALE_MAX));
     isv[RL_REWARD_SCALE_INDEX] = out;
 }