List your company

Binary deficiency detection on gold-labeled ICLR (test)

86AccuracyTADDLE (Qwen3-30B-A3B + 9B integrator)
28.914443.734758.55573.3753May 26, 2026

Evaluation results

MethodLinks
TADDLE (Qwen3-30B-A3B + 9B integrator)Family=Ours, Backbone=Qwen3-30B-A3B, Integrator=9B2026.058691.5474.7682.3
Longformer-base-4096Family=PLM2026.0583.3388.0852.8666.07
BERT-base-uncasedFamily=PLM2026.0577.3373.0839.0550.9
RottenReviews (LR)Family=Review-quality detector2026.056869.2395.2480.18
ReviewGuard (Llama-3.1-8B)Family=Review-quality detector2026.056646.9273.8157.37
RottenReviews (Llama-3-8B FT)Family=Review-quality detector, Fine-tuned=true2026.0565.3371.1582.3876.36
ReviewGuard (Qwen3-8B)Family=Review-quality detector2026.0563.3343.0857.1449.12
GLM 4.7Family=Few-shot LLM2026.0561.5673.4661.4366.91
RottenReviews (XGBoost)Family=Review-quality detector2026.0560.2274.6261.967.67
RottenReviews (Random Forest)Family=Review-quality detector2026.0558.8972.3166.1969.11
Qwen3.5-PlusFamily=Few-shot LLM2026.0555.3374.2345.7156.58
Qwen3-30BFamily=Few-shot LLM2026.0554.4468.8551.959.19
DeepSeek V3.2Family=Few-shot LLM2026.0549.7887.3124.7638.58
BigBird-RoBERTa-baseFamily=PLM2026.054633.0869.5244.83
RottenReviews (MLP)Family=Review-quality detector2026.0531.11000
RottenReviews (Qwen3-8B 0-shot)Family=Review-quality detector, Shot-count=0-shot2026.0531.11000