Merge pull request #135 from zafstojano/fix/power-function

fix(env): Power Function

reasoning_gym/arithmetic/power_function.py

@@ -7,7 +7,24 @@ from typing import Dict, Optional
 
 from ..factory import ProceduralDataset, register_dataset
 
-QUESTION_TEMPLATE = """Compute {base}^{exponent}"""
+QUESTION_TEMPLATE = """Your task is to compute an exponentiation of a number.
+
+Example:
+- Input: Compute 2^3
+- Output: 8
+- Explanation:
+    - 2^3 = 2 * 2 * 2 = 8
+    - Therefore, the final answer is 8
+
+Example:
+- Input: Compute 412.5^3
+- Output: 70189453.125
+- Explanation:
+    - 412.5^3 = 412.5 * 412.5 * 412.5 = 70189453.125
+    - Therefore, the final answer is 70189453.125
+
+Compute {base}^{exponent}
+"""
 
 
 @dataclass
@@ -32,28 +49,31 @@ class PowerFunctionDataset(ProceduralDataset):
     def score_answer(self, answer: Optional[str], entry: Dict[str, any]) -> float:
         """Overwrite this method in derived classes if a single oracle answer is not available."""
         oracle_answer = entry["answer"]
-        reward = 0.0
         if answer is not None:
-            difference = abs(float(answer) - float(oracle_answer))
+            try:
-            if difference < 1e-6:
+                answer = round(float(answer), 4)
-                reward = 1.0
+                oracle_answer = round(float(oracle_answer), 4)
-            elif difference < 1e-1:
+                difference = abs(float(answer) - float(oracle_answer))
-                reward = 0.5
+                if difference < 1e-4:
-            else:
+                    return 1.0
-                reward = 0.01
+                elif difference < 1e-1:
-
+                    return 0.5
-        return reward
+                else:
+                    return 0.01
+            except Exception as e:
+                return 0.01
+        return 0.0
 
     def __getitem__(self, idx: int) -> dict:
         """Generate a single Power Function question"""
         rng = Random(self.seed + idx)
 
-        base = rng.uniform(self.config.min_base, self.config.max_base)
+        base = round(rng.uniform(self.config.min_base, self.config.max_base), 4)
         exponent = rng.randint(self.config.min_exponent, self.config.max_exponent)
         answer = pow(base, exponent)
 
         return {
-            "question": f"Compute {base}^{exponent}",
+            "question": QUESTION_TEMPLATE.format(base=base, exponent=exponent),
             "answer": str(answer),
             "metadata": {"base": base, "exponent": exponent, "solution": answer},
         }