[evaluation] Fix xpia simulator logging warning (Azure#37351)

* Remove Optional type annotation from get() methods

* Remove duplicate overloads

* Include xpia in handled_metrics for evaluation aggregation

* Remove jailbreak_type param from logging attributes

* Rename xpia_simulator.py to indirect_attack_simulator.py'

* Remove old commits

Author: diondrapeck

sdk/evaluation/azure-ai-evaluation/azure/ai/evaluation/synthetic/__init__.py

@@ -1,6 +1,6 @@
 from .adversarial_scenario import AdversarialScenario
 from .adversarial_simulator import AdversarialSimulator
 from .direct_attack_simulator import DirectAttackSimulator
-from .xpia_simulator import IndirectAttackSimulator
+from .indirect_attack_simulator import IndirectAttackSimulator
 
 __all__ = ["AdversarialSimulator", "AdversarialScenario", "DirectAttackSimulator", "IndirectAttackSimulator"]

sdk/evaluation/azure-ai-evaluation/azure/ai/evaluation/synthetic/adversarial_simulator.py

@@ -44,15 +44,13 @@ def wrapper(*args, **kwargs):
         scenario = str(kwargs.get("scenario", None))
         max_conversation_turns = kwargs.get("max_conversation_turns", None)
         max_simulation_results = kwargs.get("max_simulation_results", None)
-        _jailbreak_type = kwargs.get("_jailbreak_type", None)
         decorated_func = monitor_operation(
             activity_name="adversarial.simulator.call",
             activity_type=ActivityType.PUBLICAPI,
             custom_dimensions={
                 "scenario": scenario,
                 "max_conversation_turns": max_conversation_turns,
                 "max_simulation_results": max_simulation_results,
-                "_jailbreak_type": _jailbreak_type,
             },
         )(func)