Question Answering on MathQA
78.7AccuracyLinUCB
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LinUCBContextual=true2026.02 | 78.7 | 8.7 | |
| LinUCBAlgorithm Family=Contextual Linear2026.02 | 78.7 | — | |
| No-RewriteAlgorithm Family=Base2026.02 | 70 | — | |
| FULLModel=Qwen3-8b, Zero-shot protocol=true2025.12 | 49.61 | — | |
| Qwen-3-4BCompression Ratio=Original2025.10 | 47 | — | |
| SPONModel=Qwen3-8b, Zero-shot protocol=true2025.12 | 46.7 | — | |
| TEALModel=Qwen3-8b, Zero-shot protocol=true2025.12 | 46.03 | — | |
| Qwen-3-8BCompression Ratio=Original2025.10 | 45 | — | |
| NeuronMLPBase Model=Qwen-3-8B, Compression Ratio=0.052025.10 | 45 | — | |
| NeuronMLPBase Model=Qwen-3-4B, Compression Ratio=0.052025.10 | 44 | — | |
| NeuronMLPBase Model=Qwen-3-8B, Compression Ratio=0.102025.10 | 44 | — | |
| NeuronMLPBase Model=Qwen-3-4B, Compression Ratio=0.102025.10 | 42 | — | |
| NeuronMLPBase Model=Qwen-3-8B, Compression Ratio=0.202025.10 | 40 | — | |
| FULLModel=Llama3-8b, Zero-shot protocol=true2025.12 | 39.46 | — | |
| Llama-3.1-8BCompression Ratio=Original2025.10 | 38 | — | |
| Qwen-3-1.7BCompression Ratio=Original2025.10 | 38 | — | |
| SPONModel=Llama3-8b, Zero-shot protocol=true2025.12 | 37.12 | — | |
| NeuronMLPBase Model=Qwen-3-1.7B, Compression Ratio=0.052025.10 | 37 | — | |
| NeuronMLPBase Model=Qwen-3-4B, Compression Ratio=0.202025.10 | 37 | — | |
| FULLModel=Mistral-v2-7B, Zero-shot protocol=true2025.12 | 36.92 | — | |
| TEALModel=Llama3-8b, Zero-shot protocol=true2025.12 | 36.11 | — | |
| NeuronMLPBase Model=Llama-3.1-8B, Compression Ratio=0.052025.10 | 36 | — | |
| Llama-3.2-3BCompression Ratio=Original2025.10 | 35 | — | |
| SPONModel=Mistral-v2-7B, Zero-shot protocol=true2025.12 | 34.81 | — | |
| TEALModel=Mistral-v2-7B, Zero-shot protocol=true2025.12 | 34.67 | — | |
| NeuronMLPBase Model=Qwen-3-1.7B, Compression Ratio=0.102025.10 | 34 | — | |
| NeuronMLPBase Model=Llama-3.2-3B, Compression Ratio=0.052025.10 | 33 | — | |
| NeuronMLPBase Model=Llama-3.1-8B, Compression Ratio=0.102025.10 | 33 | — | |
| NeuronMLPBase Model=Qwen-3-1.7B, Compression Ratio=0.202025.10 | 31 | — | |
| NeuronMLPBase Model=Llama-3.1-8B, Compression Ratio=0.202025.10 | 30 | — | |
| Llama-3.2-1BCompression Ratio=Original2025.10 | 29 | — | |
| NeuronMLPBase Model=Llama-3.2-1B, Compression Ratio=0.052025.10 | 29 | — | |
| NeuronMLPBase Model=Llama-3.2-3B, Compression Ratio=0.102025.10 | 28 | — | |
| NeuronMLPBase Model=Llama-3.2-3B, Compression Ratio=0.202025.10 | 27 | — | |
| NeuronMLPBase Model=Llama-3.2-1B, Compression Ratio=0.102025.10 | 26 | — | |
| NeuronMLPBase Model=Llama-3.2-1B, Compression Ratio=0.202025.10 | 25 | — |