List your company

Question Answering on SQuAD (Aclean, Aattack, ASR, Time)

54.52Accuracy (Clean)AGLS
19.388828.509437.6346.7506May 26, 2025

Evaluation results

MethodLinks
AGLSModels=gpt-4o-latest2025.0554.5243.4920.230.63
AGLSModels=gpt-42025.0551.8432.4337.440.78
AGLSModels=llama3.1-70B2025.0547.8327.3342.861.13
AGLSModels=gemma2-27B2025.0544.1537.7914.410.85
AGLSModels=gemma2-9B2025.0534.1127.4219.610.45
AGLSModels=llama3.1-8B2025.0533.7830.439.920.56
AGLSModels=qwen2.5-14B2025.0530.7730.431.11.12
AGLSModels=llama3.2-3B2025.0528.0922.4120.220.36
AGLSModels=qwen2.5-7B2025.0520.7419.46.460.72