List your company

Commonsense Reasoning on StoryCloze

97.6AccuracyTraditional MTL
64.517673.106381.69590.2837Mar 30, 2023Sep 19, 2023Mar 11, 2024Sep 1, 2024Feb 22, 2025Aug 15, 2025Feb 5, 2026

Evaluation results

MethodLinks
Traditional MTL2024.0597.6—
Task ArithmeticValidation=true2024.0595.3—
Individual2024.0594.9—
Fisher MergingValidation=true2024.0594.1—
EMR-MERGINGValidation=false2024.0593.6—
RegMeanValidation=true2024.0592.5—
Weight AveragingValidation=false2024.0592.5—
Ties-MergingValidation=false2024.0591.1—
Ties-MergingValidation=true2024.0590.3—
PaLM 2-Lprompting=1-shot2023.0587.4—
PaLM 2-Mprompting=1-shot2023.0586.7—
PaLMprompting=1-shot2023.0586.1—
PaLM 2-Sprompting=1-shot2023.0585.6—
SALAD-7BInput Modality=Text, Parameters=7B2025.1084.9-2
GPT-3Evaluation Protocol=1-shot2023.0384.7—
SALAD-3BInput Modality=Text, Parameters=3B2025.1082.70.2
OPT-66BEvaluation Protocol=1-shot2023.0381.83—
GLM-4-Voice-9BInput Modality=Text, Parameters=9B2025.1081.617.2
Task ArithmeticValidation=false2024.0581.5—
Qwen2-Audio-7BInput Modality=Text, Parameters=7B2025.1081.1-0.2
DiVA-Llama3.1-8BInput Modality=Text, Parameters=8B2025.1080.97.4
BloombergGPTEvaluation Protocol=1-shot2023.0380.87—
Qwen2.5-Omni-7BInput Modality=Text, Parameters=7B2025.1080.54.5
BLOOM-176BEvaluation Protocol=1-shot2023.0380.28—
GPT-NeoXEvaluation Protocol=1-shot2023.0378.3—
DSIRModel Backbone=GPT-2 XL2026.0267.72—
PPLModel Backbone=GPT-2 XL2026.0267.34—
UltraFinewebModel Backbone=GPT-2 XL2026.0267.13—
OPUSModel Backbone=GPT-2 XL2026.0267.13—
DCLM-FastTextModel Backbone=GPT-2 XL2026.0266.86—
FineWeb-EduModel Backbone=GPT-2 XL2026.0266.76—
QuRatingModel Backbone=GPT-2 XL2026.0266.7—
RandomModel Backbone=GPT-2 XL2026.0266.38—
GREATSModel Backbone=GPT-2 XL2026.0265.79—