Update README.md

README.md

@@ -54,6 +54,21 @@ output = tokenizer.decode(
 print(output)
 ```
 
+## Evaluation
+| Category                         | Task              | Llama 3.2-3B | Gemma 3-4B | Qwen2.5-3B | xGen-small 4B Instruct |
+| :------------------------------- | :---------------- | :----------- | :--------- | :--------- | :--------------------- |
+| General Knowledge & Reasoning    | MMLU              | 61.8         | 56.0       | 65.3       | 61.4                   |
+| General Knowledge & Reasoning    | MMLU-Pro          | 31.4         | 34.4       | 33.3       | 43.6                   |
+| Chat                             | Arena-Hard-v1.0   | 14.7         | 57.6       | 26.9       | 34.2                   |
+| Chat                             | MT-Bench          | 7.81         | 8.29       | 7.93       | 8.05                   |
+| Math & Science                   | GPQA              | 29.7         | 31.5       | 31.9       | 35.9                   |
+| Math & Science                   | GSM8K             | 72.3         | 90.7       | 86.3       | 92.0                   |
+| Math & Science                   | MATH              | 44.6         | 73.7       | 67.3       | 83.8                   |
+| Math & Science                   | AIME 2024         | 3.3          | 10.0       | 10.0       | 30.0                   |
+| Coding                           | HumanEval+        | 48.8         | 63.4       | 67.7       | 74.4                   |
+| Coding                           | MBPP+             | 52.1         | 66.9       | 61.9       | 52.6                   |
+| Coding                           | LiveCodeBench     | 7.4          | 13.6       | 10.1       | 32.6                   |
+
 ## Ethical Considerations
 This release is for research purposes only in support of an academic paper. Our models, datasets, and code are not specifically designed or evaluated for all downstream purposes. We strongly recommend users evaluate and address potential concerns related to accuracy, safety, and fairness before deploying this model. We encourage users to consider the common limitations of AI, comply with applicable laws, and leverage best practices when selecting use cases, particularly for high-risk scenarios where errors or misuse could significantly impact people's lives, rights, or safety. For further guidance on use cases, refer to our AUP and AI AUP.