Add evaluation results for GPQA, HLE

## Evaluation Results

This PR adds evaluation results extracted from the Model Card.

**Benchmarks:**
- GPQA: 83.8
- HLE: 12.6
- HLE: 22.29

**Files created:**
- .eval_results/gpqa.yaml
- .eval_results/hle.yaml
- .eval_results/hle_with_tools.yaml

---

Extracted automatically using the [LLM-powered evaluation extractor](https://github.com/huggingface/community-evals).

.eval_results/gpqa.yaml

@@ -0,0 +1,9 @@
+- dataset:
+    id: Idavidrein/gpqa
+    task_id: diamond
+  value: 83.8
+  date: '2026-02-13'
+  source:
+    url: https://huggingface.co/Nanbeige/Nanbeige4.1-3B
+    name: Model Card
+    user: SaylorTwift

.eval_results/hle.yaml

@@ -0,0 +1,10 @@
+- dataset:
+    id: cais/hle
+    task_id: hle
+  value: 12.6
+  date: '2026-02-13'
+  source:
+    url: https://huggingface.co/Nanbeige/Nanbeige4.1-3B
+    name: Model Card
+    user: SaylorTwift
+  note: Text-only

.eval_results/hle_with_tools.yaml

@@ -0,0 +1,10 @@
+- dataset:
+    id: cais/hle
+    task_id: hle
+  value: 22.29
+  date: '2026-02-13'
+  source:
+    url: https://huggingface.co/Nanbeige/Nanbeige4.1-3B
+    name: Model Card
+    user: SaylorTwift
+  note: with tools