List your company

Multi-turn Calibration on SciQ

4.42ECE@1MTCal
2.82613.585524.34535.1045Apr 7, 2026

Evaluation results

MethodLinks
MTCalBackbone=Gemma2-9B-it2026.044.422.293.8412.173.56
DCalBackbone=Gemma2-9B-it2026.044.6812.725.3312.795.4
MTCalBackbone=Qwen2.5-7B-Instruct2026.044.785.295.6819.974.46
MTCalBackbone=Llama3.1-8B-Instruct2026.045.396.395.6520.835.43
DCalBackbone=Qwen2.5-7B-Instruct2026.045.4614.525.720.725.58
DCalBackbone=Llama3.1-8B-Instruct2026.045.816.36.2321.816.2
SLBackbone=Llama3.1-8B-Instruct2026.047.3221.5612.0523.4310.51
PSBackbone=Qwen2.5-7B-Instruct2026.047.8910.268.0320.047.19
PSBackbone=Llama3.1-8B-Instruct2026.0410.0313.3612.521.4712.67
SLBackbone=Qwen2.5-7B-Instruct2026.0412.0924.1615.0222.6211.18
P(True)Backbone=Gemma2-9B-it2026.0414.6619.3613.8914.6412.4
VerbalBackbone=Gemma2-9B-it2026.0414.9619.8316.3910.738.65
SLBackbone=Gemma2-9B-it2026.0415.0221.0818.7416.6411.91
VerbalBackbone=Llama3.1-8B-Instruct2026.0415.1232.0920.323.3519.69
PSBackbone=Gemma2-9B-it2026.0416.7518.8117.415.4312.87
P(True)Backbone=Qwen2.5-7B-Instruct2026.0417.9421.9917.1220.5316.21
P(True)Backbone=Llama3.1-8B-Instruct2026.0418.1922.3717.9121.2717.17
VerbalBackbone=Qwen2.5-7B-Instruct2026.0420.1635.4824.3623.6321.29
SCBackbone=Qwen2.5-7B-Instruct2026.0421.3724.223.5227.818.74
SCBackbone=Llama3.1-8B-Instruct2026.0432.9830.138.7139.129.38
SCBackbone=Gemma2-9B-it2026.0444.2747.4155.5647.8138.97