feat(benchmark): support multi-dataset evaluation with configurable top-k list

- Evaluate multiple datasets in a single run (CIFAR10 + CIFAR100)
- Report Recall@K for a list of K values from one underlying search
- Save results to disk: summary.md, metrics.csv, per-dataset predictions.csv, confusion_matrix.csv
- Richer evaluation output: query_labels, topk_ids, topk_labels for downstream analysis
- Add --dataset and --top-k CLI overrides for benchmark command
- Update config schema: dataset→datasets, top_k→top_k_list

mini-nav/configs/models.py

@@ -129,7 +129,24 @@ class BenchmarkTaskConfig(BaseModel):
 
     name: str = Field(default="recall_at_k", description="Task name")
     type: str = Field(default="retrieval", description="Task type")
-    top_k: int = Field(default=10, gt=0, description="Top K for recall evaluation")
+    top_k_list: list[int] = Field(
+        default=[1, 5, 10],
+        description="Top-K values to evaluate (all derived from a single max-K search)",
+    )
+
+    @property
+    def max_k(self) -> int:
+        """Maximum K for the underlying search; all values in top_k_list <= max_k."""
+        return max(self.top_k_list) if self.top_k_list else 1
+
+    @field_validator("top_k_list", mode="after")
+    @classmethod
+    def validate_top_k_list(cls, v: list[int]) -> list[int]:
+        if not v:
+            raise ValueError("top_k_list must contain at least one value")
+        if any(k <= 0 for k in v):
+            raise ValueError("top_k_list values must be positive")
+        return sorted(set(v))
 
     # Multi-object retrieval specific settings
     gamma: float = Field(
@@ -148,7 +165,10 @@ class BenchmarkConfig(BaseModel):
 
     model_config = ConfigDict(extra="ignore")
 
-    dataset: DatasetSourceConfig = Field(default_factory=DatasetSourceConfig)
+    datasets: list[DatasetSourceConfig] = Field(
+        default_factory=lambda: [DatasetSourceConfig()],
+        description="Dataset configurations to evaluate (supports multiple).",
+    )
     task: BenchmarkTaskConfig = Field(default_factory=BenchmarkTaskConfig)
     batch_size: int = Field(default=64, gt=0, description="Batch size for DataLoader")
     model_table_prefix: str = Field(