Question Answering on CSQA (ECE, BS)
1.57ECECaliDist (As.)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CaliDist (As.)LLM=QWEN-3 8B2026.06 | 1.57 | 12.45 | |
| CaliDist (Pr.)LLM=QWEN-3 8B2026.06 | 2.56 | 12.18 | |
| CaliDist (Sa.)LLM=LLAMA-3.1 8B2026.06 | 4.32 | 15.01 | |
| CaliDist (As.)LLM=LLAMA-3.1 8B2026.06 | 4.73 | 14.7 | |
| CaliDist (Sa.)LLM=QWEN-3 8B2026.06 | 4.8 | 13.62 | |
| CaliDist (Pr.)LLM=LLAMA-3.1 8B2026.06 | 5.32 | 14.77 | |
| Temperature ScalingLLM=LLAMA-3.1 8B2026.06 | 5.57 | 14.73 | |
| VanillaLLM=LLAMA-3.1 8B2026.06 | 12.59 | 16.57 | |
| Temperature ScalingLLM=QWEN-3 8B2026.06 | 14.83 | 15.86 | |
| EntropyLLM=QWEN-3 8B2026.06 | 14.86 | 30.4 | |
| ConsistencyLLM=QWEN-3 8B2026.06 | 14.87 | 30.37 | |
| FSDLLM=LLAMA-3.1 8B2026.06 | 15.95 | 18.89 | |
| VanillaLLM=QWEN-3 8B2026.06 | 16.39 | 16.91 | |
| FSDLLM=QWEN-3 8B2026.06 | 17.17 | 17.36 | |
| ConsistencyLLM=LLAMA-3.1 8B2026.06 | 17.87 | 19.41 | |
| EntropyLLM=LLAMA-3.1 8B2026.06 | 18.37 | 19.71 |