Mathematical Reasoning on AQuA (ECE and BS)
2.02ECECaliDist (As.)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CaliDist (As.)LLM=LLAMA-3.1 8B2026.06 | 2.02 | 17.65 | |
| CaliDist (Sa.)LLM=LLAMA-3.1 8B2026.06 | 2.73 | 17.25 | |
| CaliDist (Pr.)LLM=QWEN-3 8B2026.06 | 2.82 | 11.21 | |
| CaliDist (Pr.)LLM=LLAMA-3.1 8B2026.06 | 2.98 | 17.45 | |
| CaliDist (As.)LLM=QWEN-3 8B2026.06 | 3.27 | 11.17 | |
| CaliDist (Sa.)LLM=QWEN-3 8B2026.06 | 3.9 | 11.16 | |
| ConsistencyLLM=LLAMA-3.1 8B2026.06 | 4.88 | 14.51 | |
| FSDLLM=QWEN-3 8B2026.06 | 7.37 | 9.48 | |
| Temperature ScalingLLM=QWEN-3 8B2026.06 | 8.28 | 10.63 | |
| Temperature ScalingLLM=LLAMA-3.1 8B2026.06 | 8.71 | 17.11 | |
| EntropyLLM=QWEN-3 8B2026.06 | 8.9 | 10.86 | |
| ConsistencyLLM=QWEN-3 8B2026.06 | 9.15 | 9.63 | |
| VanillaLLM=QWEN-3 8B2026.06 | 12.54 | 12.7 | |
| CaliDist_verbalized (As.)LLM=QWEN-2 7B2026.06 | 13.71 | 17.68 | |
| CaliDist_verbalized (Sa.)LLM=QWEN-2 7B2026.06 | 14.06 | 17.41 | |
| CaliDist_verbalized (Pr.)LLM=QWEN-2 7B2026.06 | 14.44 | 17.83 | |
| VerbalizedLLM=QWEN-2 7B2026.06 | 16.74 | 17.63 | |
| FSDLLM=LLAMA-3.1 8B2026.06 | 20.07 | 19.63 | |
| CaliDist_verbalized (Pr.)LLM=LLAMA-3.1 8B2026.06 | 20.16 | 25.04 | |
| VerbalizedLLM=LLAMA-3.1 8B2026.06 | 20.57 | 24.98 | |
| CaliDist_verbalized (As.)LLM=LLAMA-3.1 8B2026.06 | 20.59 | 26.46 | |
| CaliDist_verbalized (Sa.)LLM=LLAMA-3.1 8B2026.06 | 20.84 | 24.16 | |
| VanillaLLM=LLAMA-3.1 8B2026.06 | 23.18 | 22.82 | |
| EntropyLLM=LLAMA-3.1 8B2026.06 | 33.22 | 28.2 |