Question Answering on AdversarialQA (val)
38.5EMT5-base
Evaluation results
| Method | Links | ||
|---|---|---|---|
| T5-baseModel Size=220M, Data=Full, Method=Fine-Tuned2024.03 | 38.5 | — | |
| GOLDBase Model=LLaMA2 -> T5-base, Model Size=220M, Data=3K2024.03 | 25.5 | — | |
| T5-baseModel Size=220M, Data=Full, Method=Pre-Trained2024.03 | 25.3 | — | |
| LLaMA2Model Size=7B, Method=Few-shot, Prompt=Specific2024.03 | 25.1 | — | |
| P2ModelBase Model=LLaMA2 -> T5-base, Model Size=220M, Data=3K2024.03 | 25 | — | |
| ZeroGenBase Model=LLaMA2 -> T5-base, Model Size=220M, Data=3K2024.03 | 21.3 | — | |
| ProGenBase Model=LLaMA2 -> T5-base, Model Size=220M, Data=3K2024.03 | 20.5 | — | |
| DistilBERTPLM=Gold Data, Setting=SUPERVISED, #Param=66M2022.02 | 18.6 | 29.85 | |
| LLaMA2Model Size=7B, Method=Few-shot, Prompt=General2024.03 | 14.6 | — | |
| ZeroGenPLM=GPT2-XL, TAM=DistilBERT, Setting=ZEROGEN, #Param=66M2022.02 | 6.33 | 9.96 | |
| ZeroGenPLM=GPT2-Large, TAM=DistilBERT, Setting=ZEROGEN, #Param=66M2022.02 | 5.93 | 9.63 | |
| LSTMPLM=Gold Data, Setting=SUPERVISED, #Param=~7M2022.02 | 5.37 | 11.86 | |
| ZeroGenPLM=GPT2, TAM=DistilBERT, Setting=ZEROGEN, #Param=66M2022.02 | 5.2 | 8.26 | |
| ZeroGenPLM=GPT2-XL, TAM=LSTM, Setting=ZEROGEN, #Param=~7M2022.02 | 3.23 | 6.34 | |
| ZeroGenPLM=GPT2-Large, TAM=LSTM, Setting=ZEROGEN, #Param=~7M2022.02 | 2.33 | 5.24 | |
| GPT2-XLSetting=PROMPTING, #Param=1.5B2022.02 | 2.13 | 6.3 | |
| GPT2-LargeSetting=PROMPTING, #Param=762M2022.02 | 1.47 | 5.16 | |
| ZeroGenPLM=GPT2, TAM=LSTM, Setting=ZEROGEN, #Param=~7M2022.02 | 1 | 3.83 | |
| GPT2Setting=PROMPTING, #Param=117M2022.02 | 0.37 | 2.58 |