Classification on CSQA
2.21Expected Calibration Error (ECE)CaliDist (As.)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CaliDist (As.)LLM=Phi-4-mini (3.8B)2026.06 | 2.21 | 16.49 | |
| CaliDist (As.)LLM=Gemma-3 (4B)2026.06 | 2.32 | 18.75 | |
| CaliDist (Sa.)LLM=Gemma-3 (4B)2026.06 | 2.51 | 18.32 | |
| CaliDist (Pr.)LLM=Gemma-3 (4B)2026.06 | 2.55 | 17.36 | |
| CaliDist (Pr.)LLM=Phi-4-mini (3.8B)2026.06 | 2.89 | 16.62 | |
| CaliDist (Sa.)LLM=Phi-4-mini (3.8B)2026.06 | 3.35 | 16.58 | |
| Temperature ScalingLLM=Phi-4-mini (3.8B)2026.06 | 4.57 | 16.38 | |
| VanillaLLM=Phi-4-mini (3.8B)2026.06 | 11.13 | 17.73 | |
| ConsistencyLLM=Phi-4-mini (3.8B)2026.06 | 11.52 | 25.67 | |
| FSDLLM=Phi-4-mini (3.8B)2026.06 | 14.65 | 20.19 | |
| EntropyLLM=Phi-4-mini (3.8B)2026.06 | 20.46 | 30.34 | |
| Temperature ScalingLLM=Gemma-3 (4B)2026.06 | 23.87 | 24.18 | |
| VanillaLLM=Gemma-3 (4B)2026.06 | 25.44 | 25.43 | |
| EntropyLLM=Gemma-3 (4B)2026.06 | 25.48 | 25.53 | |
| FSDLLM=Gemma-3 (4B)2026.06 | 25.49 | 25.53 | |
| ConsistencyLLM=Gemma-3 (4B)2026.06 | 25.54 | 25.58 |