Automated MNLP evaluation report (2026-05-19)

EVAL_REPORT.md

@@ -0,0 +1,61 @@
+# Automated MNLP evaluation report
+
+- **Model repo:** [`cs-552-2026-ma-que/general_knowledge_model`](https://huggingface.co/cs-552-2026-ma-que/general_knowledge_model)
+- **Owner(s):** group **ma-que**
+- **Generated at:** 2026-05-19T03:22:52+00:00 (UTC)
+- **Pipeline:** [mnlp-project-ci](https://github.com/eric11eca/mnlp-project-ci)
+
+_This PR is opened automatically by the course CI. It is **non-blocking** — you do not need to merge it. The next nightly run will refresh this file._
+
+## Summary
+
+| Benchmark | Accuracy | Status |
+|---|---:|---|
+| Math | — | not run |
+| Knowledge | 0.0400 | ok |
+| Multilingual | — | not run |
+| Safety | — | not run |
+
+## Sample completions
+
+_Prompts are intentionally omitted to avoid revealing benchmark contents. For multi-completion problems, only one completion is shown per sample._
+
+### Knowledge
+
+**Correct** (1 shown)
+
+- **reference**: `C`
+- **overall** (1/1 completions correct)
+- **extracted** (✓): `C`
+- **completion**:
+
+    ```text
+    <think>
+    
+    </think>
+    
+    </think>
+    
+    C. A = 4-isopropylcyclohexan-1-one, B = 5-methylhexane-2-one
+    
+    Final Answer:
+    \boxed{C}
+    ```
+
+**Incorrect** (1 shown)
+
+- **reference**: `A`
+- **overall** (0/1 completions correct)
+- **extracted** (✗): ``
+- **completion**:
+
+    ```text
+    <think>
+    
+    </think>
+    
+    </think>
+    
+    Instruction:
+    Reason briefly, then put the final answer in \boxed{}.
+    ```