diff --git a/evaluate_next_event.py b/evaluate_next_event.py
index d8c8a89..7858366 100644
--- a/evaluate_next_event.py
+++ b/evaluate_next_event.py
@@ -186,6 +186,22 @@ def _compute_next_event_auc_clean_control(
 
 def main() -> None:
     args = parse_args()
+
+    # Best-effort control of implicit parallelism to avoid CPU oversubscription.
+    # Note: environment variables are ideally set before importing NumPy/PyTorch,
+    # but setting them early in main can still affect subprocesses or lazy readers.
+    if int(args.max_cpu_cores) > 0:
+        n_threads = int(args.max_cpu_cores)
+        torch.set_num_threads(n_threads)
+        for k in (
+            "OMP_NUM_THREADS",
+            "MKL_NUM_THREADS",
+            "OPENBLAS_NUM_THREADS",
+            "VECLIB_MAXIMUM_THREADS",
+            "NUMEXPR_NUM_THREADS",
+        ):
+            os.environ[k] = str(n_threads)
+        print(f"Restricting implicit parallelism to {n_threads} threads.")
     seed_everything(args.seed)
 
     show_progress = (not args.no_tqdm)