Classification on MSciNLI, MNLI, PPDB, Yahoo, HellaSwag, CSQA, AQUA Average
4.34ECECaliDist (As.)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| CaliDist (As.)LLM=Phi-4-mini (3.8B)2026.06 | 4.34 | 18.13 | |
| CaliDist (Sa.)LLM=Gemma-3 (4B)2026.06 | 4.4 | 20.18 | |
| CaliDist (As.)LLM=Gemma-3 (4B)2026.06 | 4.56 | 21.01 | |
| CaliDist (Pr.)LLM=Phi-4-mini (3.8B)2026.06 | 5.01 | 18.03 | |
| CaliDist (Sa.)LLM=Phi-4-mini (3.8B)2026.06 | 5.23 | 17.88 | |
| CaliDist (Pr.)LLM=Gemma-3 (4B)2026.06 | 5.33 | 20.4 | |
| Temperature ScalingLLM=Phi-4-mini (3.8B)2026.06 | 8.07 | 18.75 | |
| ConsistencyLLM=Phi-4-mini (3.8B)2026.06 | 16 | 22.58 | |
| VanillaLLM=Phi-4-mini (3.8B)2026.06 | 16.2 | 21.13 | |
| FSDLLM=Phi-4-mini (3.8B)2026.06 | 17.06 | 22.36 | |
| EntropyLLM=Phi-4-mini (3.8B)2026.06 | 23.65 | 26.71 | |
| Temperature ScalingLLM=Gemma-3 (4B)2026.06 | 32.42 | 33.26 | |
| ConsistencyLLM=Gemma-3 (4B)2026.06 | 34.26 | 34.96 | |
| FSDLLM=Gemma-3 (4B)2026.06 | 34.35 | 34.91 | |
| VanillaLLM=Gemma-3 (4B)2026.06 | 35.11 | 35.07 | |
| EntropyLLM=Gemma-3 (4B)2026.06 | 35.44 | 35.48 |