List your company

Short-Answer Factuality on TruthfulQA (test)

47.47MC1 Factuality ScoreSLED
31.017235.288639.5643.8314Nov 1, 2024

Evaluation results

MethodLinks
SLEDBackbone=Gemma-3-27B-IT, Decoding=SLED2024.1147.4773.5843.53—
SLEDBackbone=Gemma-3-12B-IT, Decoding=SLED2024.1146.3371.1540.9—
SLEDBackbone=Gemma-3-4B-IT, Decoding=SLED2024.1145.1966.1538.06—
Gemma-3-27B-IT (Greedy)Backbone=Gemma-3-27B-IT, Decoding=Greedy2024.1141.1464.0638.71—
Gemma-3-12B-IT (Greedy)Backbone=Gemma-3-12B-IT, Decoding=Greedy2024.1139.4963.6237.89—
SLEDBackbone=Gemma-3-27B-PT, Decoding=SLED2024.1138.4867.435.65—
SLEDBackbone=Gemma-3-12B-PT, Decoding=SLED2024.1137.8567.0534.59—
Gemma-3-4B-IT (Greedy)Backbone=Gemma-3-4B-IT, Decoding=Greedy2024.1136.9658.4234.13—
SLEDBackbone=Gemma-3-1B-IT, Decoding=SLED2024.1136.7171.7942—
DoLaBackbone=Gemma-3-1B-IT, Decoding=DoLa2024.1136.4668.8635.69—
Gemma-3-27B-PT (Greedy)Backbone=Gemma-3-27B-PT, Decoding=Greedy2024.1136.261.3934.95—
DoLaBackbone=Gemma-3-4B-IT, Decoding=DoLa2024.1136.0868.5935.58—
Gemma-3-12B-PT (Greedy)Backbone=Gemma-3-12B-PT, Decoding=Greedy2024.1135.8263.734.83—
SLEDBackbone=Gemma-3-1B-PT, Decoding=SLED2024.1135.5771.7940.69—
DoLaBackbone=Gemma-3-1B-PT, Decoding=DoLa2024.1135.4468.935.29—
SLEDBackbone=Gemma-3-4B-PT, Decoding=SLED2024.1135.1969.2538.96—
DoLaBackbone=Gemma-3-12B-IT, Decoding=DoLa2024.1135.1967.5335.04—
DoLaBackbone=Gemma-3-4B-PT, Decoding=DoLa2024.1134.6868.2134.9—
Gemma-3-4B-PT (Greedy)Backbone=Gemma-3-4B-PT, Decoding=Greedy2024.1133.864.1232.64—
DoLaBackbone=Gemma-3-12B-PT, Decoding=DoLa2024.1133.5466.6234.25—
Gemma-3-1B-IT (Greedy)Backbone=Gemma-3-1B-IT, Decoding=Greedy2024.1133.2961.6530.74—
DoLaBackbone=Gemma-3-27B-IT, Decoding=DoLa2024.1133.1665.9333.63—
Gemma-3-1B-PT (Greedy)Backbone=Gemma-3-1B-PT, Decoding=Greedy2024.1132.2863.2230.98—
DoLaBackbone=Gemma-3-27B-PT, Decoding=DoLa2024.1131.6565.0332.72—
STARLING (BLUE-RED)Data strategy=BLUE-RED2023.08———49.6
STARLING (BLUE)Data strategy=BLUE2023.08———48.9
VICUNA-7BModel size=7B2023.08———46.99
VICUNA-FT-7BModel size=7B, Training=Fine-tuned2023.08———48.85