Update simulator-interaction-data.md

Author: LHL407

articles/ai-foundry/how-to/develop/simulator-interaction-data.md

@@ -180,7 +180,7 @@ The `Simulator` class offers extensive customization options. With these options
 
 #### Query and response generation Prompty customization
 
-The `query_response_generating_prompty_override` allows you to customize how query-response pairs are generated from input text. This capability is useful when you want to control the format or content of the generated responses as input to your simulator.
+The `query_response_generating_prompty_override` parameter allows you to customize how query-response pairs are generated from input text. This capability is useful when you want to control the format or content of the generated responses as input to your simulator.
 
 ```python
 current_dir = os.path.dirname(__file__)
@@ -209,7 +209,7 @@ for output in outputs:
 
 #### Simulation Prompty customization
 
-The `Simulator` class uses a default Prompty that instructs the LLM on how to simulate a user interacting with your application. The `user_simulating_prompty_override` enables you to override the default behavior of the simulator. By adjusting these parameters, you can tune the simulator to produce responses that align with your specific requirements, enhancing the realism and variability of the simulations.
+The `Simulator` class uses a default Prompty that instructs the LLM on how to simulate a user interacting with your application. The `user_simulating_prompty_override` parameter enables you to override the default behavior of the simulator. By adjusting these parameters, you can tune the simulator to produce responses that align with your specific requirements, enhancing the realism and variability of the simulations.
 
 ```python
 user_simulator_prompty_kwargs = {
@@ -324,7 +324,7 @@ azure_ai_project = {
 
 ### Specify target callback to simulate against for adversarial simulator
 
-You can bring any application endpoint to the adversarial simulator. The `AdversarialSimulator` class supports sending service-hosted queries and receiving responses with a callback function, as defined in the following code block. The `AdversarialSimulator` adheres to the [OpenAI messages protocol](https://platform.openai.com/docs/api-reference/messages/object#messages/object-content).
+You can bring any application endpoint to the adversarial simulator. The `AdversarialSimulator` class supports sending service-hosted queries and receiving responses with a callback function, as defined in the following code block. The `AdversarialSimulator` class adheres to the [OpenAI messages protocol](https://platform.openai.com/docs/api-reference/messages/object#messages/object-content).
 
 ```python
 async def callback(
@@ -385,7 +385,7 @@ By default, we run simulations asynchronously. We enable optional parameters:
 
 ## Supported adversarial simulation scenarios
 
-The `AdversarialSimulator` supports a range of scenarios, hosted in the service, to simulate against your target application or function:
+The `AdversarialSimulator` class supports a range of scenarios, hosted in the service, to simulate against your target application or function:
 
 | Scenario                  | Scenario enumeration                | Maximum number of simulations | Use this dataset for evaluating |
 |-------------------------------|------------------------------|---------|---------------------|
@@ -408,7 +408,7 @@ Evaluating vulnerability toward the following types of jailbreak attacks is supp
 - **Direct attack jailbreak**: This type of attack, also known as a user prompt injected attack (UPIA), injects prompts in the user role turn of conversations or queries to generative AI applications.
 - **Indirect attack jailbreak**: This type of attack, also known as a cross domain prompt injected attack (XPIA), injects prompts in the returned documents or context of the user's query to generative AI applications.
 
-*Evaluating direct attack* is a comparative measurement that uses the Azure AI Content Safety evaluators as a control. It isn't its own AI-assisted metric. Run `ContentSafetyEvaluator` on two different, red-teamed datasets generated by `AdversarialSimulator`:
+*Evaluating direct attack* is a comparative measurement that uses the Azure AI Content Safety evaluators as a control. It isn't its own AI-assisted metric. Run `ContentSafetyEvaluator` on two different, red-teamed datasets generated by the `AdversarialSimulator` class:
 
 - Baseline adversarial test dataset using one of the previous scenario enumerations for evaluating hateful and unfair content, sexual content, violent content, and self-harm-related content
 - Adversarial test dataset with direct attack jailbreak injections in the first turn:
@@ -431,7 +431,7 @@ The outputs consist of two lists:
 
 Run two evaluation runs with `ContentSafetyEvaluator` and measure the differences between the two datasets' defect rates.
 
-*Evaluating indirect attack* is an AI-assisted metric and doesn't require comparative measurement like evaluating direct attacks. You can generate an indirect attack jailbreak-injected dataset with the following, and then evaluate with the `IndirectAttackEvaluator`.
+*Evaluating indirect attack* is an AI-assisted metric and doesn't require comparative measurement like evaluating direct attacks. You can generate an indirect attack jailbreak-injected dataset with the following, and then evaluate with `IndirectAttackEvaluator`.
 
 ```python
 indirect_attack_simulator=IndirectAttackSimulator(azure_ai_project=azure_ai_project, credential=credential)
@@ -445,9 +445,9 @@ outputs = await indirect_attack_simulator(
 
 ### Output
 
-The `output` is a `JSON` array of messages and adheres to the OpenAI messages protocol. You can learn more [in this OpenAI resource](https://platform.openai.com/docs/api-reference/messages/object#messages/object-content).
+The output is a JSON array of messages and adheres to the OpenAI messages protocol. You can learn more [in this OpenAI resource](https://platform.openai.com/docs/api-reference/messages/object#messages/object-content).
 
-The `messages` in `output` is a list of role-based turns. For each turn, it contains the following elements:
+The `messages` output is a list of role-based turns. For each turn, it contains the following elements:
 
 - `content`: The content of an interaction.
 - `role`: Either the user (simulated agent) or assistant, and any required citations or context from either the simulated user or the chat application.
@@ -506,7 +506,7 @@ For single-turn simulations, use the helper function `to_eval_qr_json_lines()` t
 
 #### Multi-language adversarial simulation
 
-The `AdversarialSimulator` uses the [ISO standard](https://www.andiamo.co.uk/resources/iso-language-codes/) and supports the following languages:
+The `AdversarialSimulator` class uses the [ISO standard](https://www.andiamo.co.uk/resources/iso-language-codes/) and supports the following languages:
 
 | Language           | ISO language code |
 |--------------------|-------------------|
@@ -534,7 +534,7 @@ outputs = await simulator(
 
 #### Set the randomization seed
 
-By default, the `AdversarialSimulator` randomizes interactions in every simulation. You can set a `randomization_seed` parameter to produce the same set of conversation starters every time for reproducibility.
+By default, the `AdversarialSimulator` class randomizes interactions in every simulation. You can set a `randomization_seed` parameter to produce the same set of conversation starters every time for reproducibility.
 
 ```python
 outputs = await simulator(
@@ -566,7 +566,7 @@ This function can stop a conversation if the conversation meets certain criteria
 
 The scenario simulator supports retry logic. The default maximum number of retries in case the last API call failed is 3. The default number of seconds to sleep between consequent retries in case the last API call failed is 3.
 
-Users can also define their own `api_call_retry_sleep_sec` and `api_call_retry_max_count` and pass it in while running the function call in `simulate()`.
+Users can also define their own `api_call_retry_sleep_sec` and `api_call_retry_max_count` values and pass the values in while running the function call in `simulate()`.
 
 ## Related content