Improve all-future first-onset training

eval_data.py

@@ -66,7 +66,6 @@ def validate_training_mode_config(cfg: Dict[str, Any]) -> None:
             "model_target_mode must be next_token or all_future, got "
             f"{model_target_mode!r}"
         )
-
     disease_history_mode = normalize_disease_history_mode(
         cfg.get("disease_history_mode", DISEASE_HISTORY_MODE_TIMED)
     )
@@ -183,6 +182,11 @@ def build_model_from_dataset(
             "model_target_mode must be next_token or all_future, got "
             f"{model_target_mode!r}"
         )
+    risk_head_bias = bool(cfg_get(args, cfg, "risk_head_bias", False))
+    if state_dict is not None:
+        # The checkpoint schema is authoritative. This keeps all older
+        # bias-free checkpoints loadable while restoring the new baseline bias.
+        risk_head_bias = "risk_head.bias" in state_dict
     model_architecture = resolve_model_architecture(cfg, state_dict)
     continuous_value_center = None
     continuous_value_scale = None
@@ -231,6 +235,7 @@ def build_model_from_dataset(
         dist_mode=str(cfg_get(args, cfg, "dist_mode", "exponential")),
         dropout=float(cfg_get(args, cfg, "dropout", 0.0)),
         model_architecture=model_architecture,
+        risk_head_bias=risk_head_bias,
     )