List your company

Mathematical Reasoning on GSM8K (ECE, BS)

4.28ECECaliDist (As.)
3.6068.155512.70517.2545Jun 4, 2026

Evaluation results

MethodLinks
CaliDist (As.)Backbone=QWEN-3 8B2026.064.287.16
CaliDist (As.)Backbone=LLAMA-3.1 8B2026.065.5114.13
ConsistencyBackbone=LLAMA-3.1 8B2026.066.127.83
FSDBackbone=QWEN-3 8B2026.066.186.53
EntropyBackbone=QWEN-3 8B2026.066.786.75
ConsistencyBackbone=QWEN-3 8B2026.066.976.91
VanillaBackbone=QWEN-3 8B2026.0610.4810.48
FSDBackbone=LLAMA-3.1 8B2026.0613.1911.73
VanillaBackbone=LLAMA-3.1 8B2026.0616.7116.8
EntropyBackbone=LLAMA-3.1 8B2026.0621.1316.7