Test strategy for dictator game

README.md

@@ -19,6 +19,9 @@ The figure below presents a violin plot depicting the share of the total amount
 that the dictator allocates to themselves for each model. 
 The temperature is fixed at 0.7, and each experiment was conducted 30 times.
 The median share taken by GPT-4.5, Llama3, Mistral-Small, and DeepSeek-R1 is 50.
+When we prompt the models to generate a strategy in the form of an algorithm implemented 
+in the Python programming language, rather than generating an action, all models divide 
+the amount fairly except for GPT-4.5, which takes approximately 70% of the total amount for itself.
 It is worth noticing that, under these standard conditions, humans typically keep an average of around \$80
 (Fortsythe et al. 1994). It is interesting to note that the variability observed between different executions 
 in the responses of the same LLM is comparable to the diversity of behaviors observed in humans. In other words, 

src/dictator/dictator.py

@@ -21,34 +21,39 @@ class AgentResponse(BaseModel):
 
 # The dictator game simulation class
 class Dictator:
-    def __init__(self, amount: int, model: str, temperature: float, max_retries: int = 3):
+    def __init__(self, amount: int, model: str, temperature: float, strategy=False, max_retries: int = 3):
         self.amount = amount
         self.model = model
         self.temperature = temperature
+        self.strategy = strategy
         self.max_retries = max_retries  # Maximum retry attempts in case of hallucinations
 
-        is_openai_model = model.startswith("gpt")
-        base_url = "https://api.openai.com/v1" if is_openai_model else "http://localhost:11434/v1"
+        if not strategy:
+            is_openai_model = model.startswith("gpt")
+            base_url = "https://api.openai.com/v1" if is_openai_model else "http://localhost:11434/v1"
 
-        model_info = {
-            "temperature": self.temperature,
-            "function_calling": True,
-            "parallel_tool_calls": True,
-            "family": "unknown",
-            "json_output": True,
-            "vision": False
-        }
+            model_info = {
+                "temperature": self.temperature,
+                "function_calling": True,
+                "parallel_tool_calls": True,
+                "family": "unknown",
+                "json_output": True,
+                "vision": False
+            }
 
-        self.model_client = OpenAIChatCompletionClient(
-            model=self.model,
-            base_url=base_url,
-            api_key=OPENAI_API_KEY,
-            model_info=model_info,
-            response_format=AgentResponse
-        )
+            self.model_client = OpenAIChatCompletionClient(
+                model=self.model,
+                base_url=base_url,
+                api_key=OPENAI_API_KEY,
+                model_info=model_info,
+                response_format=AgentResponse
+            )
 
     async def run(self) -> Dict:
-        """Runs the model and ensures a valid response."""
+        """Runs the model if strategy is False, otherwise uses a classical method."""
+        if self.strategy:
+            return self.apply_strategy()
+
         instruction = f"""
         Hi, let’s play a game. You are paired with another player.
         Your role is to decide how to divide ${self.amount}, and the other player simply receives your choice.
@@ -84,8 +89,39 @@ class Dictator:
 
         raise ValueError("Model failed to provide a valid response after multiple attempts.")
 
+    def apply_strategy(self) -> Dict:
+        """Generates a response."""
+        if (self.model == "gpt-4.5-preview-2025-02-27") :
+            my_share = int(0.7 * self.amount)  # Example rule: keep 70%
+            other_share = self.amount - my_share
+            motivation = f"Using strategy from {self.model}, I chose to keep {my_share} and give {other_share}."
+            return {"my_share": my_share, "other_share": other_share, "motivation": motivation}
+        if (self.model == "llama3") :
+            my_share = self.amount / 2
+            other_share = self.amount - my_share
+            motivation = "I'm being fair and generous!"
+            agent_response = AgentResponse(my_share=my_share, other_share=other_share, motivation=motivation)
+            return agent_response.model_dump()
+        if (self.model == "mistral-small") :
+            my_share = self.amount // 2
+            other_share = self.amount - my_share
+            motivation = "The decision is to divide the money equally."
+            agent_response = AgentResponse(
+                my_share=my_share,
+                other_share=other_share,
+                motivation=motivation
+            )
+            return agent_response
+        if (self.model == "deepseek-r1") :
+            half_amount = self.amount // 2
+            return {
+                "my_share": half_amount,
+                "other_share": half_amount,
+                "motivation": "Split equally between both players."
+            }
+
 # Run the async function and return the response
 if __name__ == "__main__":
-    game_agent = Dictator(amount=100, model="gpt-4.5-preview-2025-02-27", temperature=0.7)
-    response_json = asyncio.run(game_agent.get_valid_response())
-    print(response_json)
\ No newline at end of file
+    game_agent = Dictator(amount=100, model="laama3", temperature=0.7, strategy=True)  # Toggle strategy here
+    response_json = asyncio.run(game_agent.run())
+    print(response_json)

src/dictator/dictator_draw_violin.py

@@ -25,12 +25,24 @@ sns.violinplot(
     density_norm="width"  # Normalizes the width of the violins for comparison
 )
 
+# Add vertical lines for strategies
+strategy_values = {
+    'gpt-4.5-preview-2025-02-27': 70,
+    'llama3': 50,
+    'mistral-small': 50,
+    'deepseek-r1': 50
+}
+
+for model, value in strategy_values.items():
+    plt.axhline(y=value, color=color_palette[model], linestyle="dashed", linewidth=2, label=f"{model} strategy")
+
 # Set the y-axis limits between 0 and 100
 plt.ylim(0, 100)
 # Labels and title
 plt.xlabel("Model")
 plt.ylabel("Share of money assigned to oneself")
 plt.title("Distribution of personal share by model in the dictator game")
+plt.legend()
 
 # Save and display
 plt.savefig("../../figures/dictator/dictator_violin.svg", format="svg")
\ No newline at end of file

src/dictator/dictator_experiments.py

@@ -41,7 +41,7 @@ class DictatorExperiment:
 
 # Running the experiment
 if __name__ == "__main__":
-    models = ["gpt-4.5-preview-2025-02-27", "llama3", "mistral-small", "deepseek-r1"]
+    models = ["llama3", "mistral-small", "deepseek-r1"] # "gpt-4.5-preview-2025-02-27"
     temperature = 0.7
     amount = 100
     iterations = 30