Question Answering on 5 QA tasks
54.02AccuracyKLEfull
Evaluation Results
| Method | Links | ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| KLEfullBackbone=Mistral-Nemo, Access=White-box2025.05 | 54.02 | 0.75 | — | — | — | — | — | — | — | — | — | — | — | — | |
| SEBackbone=Mistral-Nemo, Access=White-box2025.05 | 54.02 | — | 0.72 | — | — | — | — | — | — | — | — | — | — | — | |
| NEBackbone=Mistral-Nemo, Access=White-box2025.05 | 54.02 | — | — | 0.7 | — | — | — | — | — | — | — | — | — | — | |
| WSNNEBackbone=Mistral-Nemo, Access=White-box2025.05 | 54.02 | — | — | — | 0.74 | — | — | — | — | — | — | — | — | — | |
| KLEheatBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | 0.72 | — | — | — | — | — | — | — | — | |
| DSEBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | — | 0.72 | — | — | — | — | — | — | — | |
| pTrueBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | — | — | 0.7 | — | — | — | — | — | — | |
| NumSetBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | — | — | — | 0.72 | — | — | — | — | — | |
| LexSimBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | — | — | — | — | 0.72 | — | — | — | — | |
| SumEigvBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | — | — | — | — | — | 0.72 | — | — | — | |
| DegBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | — | — | — | — | — | — | 0.72 | — | — | |
| EccenBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | — | — | — | — | — | — | — | 0.73 | — | |
| SNNEBackbone=Mistral-Nemo, Access=Black-box2025.05 | 54.02 | — | — | — | — | — | — | — | — | — | — | — | — | 0.73 | |
| KLEfullBackbone=Llama-3.1-8B, Access=White-box2025.05 | 50.76 | 0.73 | — | — | — | — | — | — | — | — | — | — | — | — | |
| SEBackbone=Llama-3.1-8B, Access=White-box2025.05 | 50.76 | — | 0.71 | — | — | — | — | — | — | — | — | — | — | — | |
| NEBackbone=Llama-3.1-8B, Access=White-box2025.05 | 50.76 | — | — | 0.72 | — | — | — | — | — | — | — | — | — | — | |
| WSNNEBackbone=Llama-3.1-8B, Access=White-box2025.05 | 50.76 | — | — | — | 0.74 | — | — | — | — | — | — | — | — | — | |
| KLEheatBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | 0.71 | — | — | — | — | — | — | — | — | |
| DSEBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | — | 0.71 | — | — | — | — | — | — | — | |
| pTrueBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | — | — | 0.66 | — | — | — | — | — | — | |
| NumSetBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | — | — | — | 0.7 | — | — | — | — | — | |
| LexSimBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | — | — | — | — | 0.71 | — | — | — | — | |
| SumEigvBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | — | — | — | — | — | 0.71 | — | — | — | |
| DegBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | — | — | — | — | — | — | 0.71 | — | — | |
| EccenBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | — | — | — | — | — | — | — | 0.71 | — | |
| SNNEBackbone=Llama-3.1-8B, Access=Black-box2025.05 | 50.76 | — | — | — | — | — | — | — | — | — | — | — | — | 0.74 | |
| KLEfullBackbone=Llama2-13B, Access=White-box2025.05 | 46.07 | 0.67 | — | — | — | — | — | — | — | — | — | — | — | — | |
| SEBackbone=Llama2-13B, Access=White-box2025.05 | 46.07 | — | 0.65 | — | — | — | — | — | — | — | — | — | — | — | |
| NEBackbone=Llama2-13B, Access=White-box2025.05 | 46.07 | — | — | 0.63 | — | — | — | — | — | — | — | — | — | — | |
| WSNNEBackbone=Llama2-13B, Access=White-box2025.05 | 46.07 | — | — | — | 0.66 | — | — | — | — | — | — | — | — | — | |
| KLEheatBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | 0.65 | — | — | — | — | — | — | — | — | |
| DSEBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | — | 0.64 | — | — | — | — | — | — | — | |
| pTrueBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | — | — | 0.63 | — | — | — | — | — | — | |
| NumSetBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | — | — | — | 0.63 | — | — | — | — | — | |
| LexSimBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | — | — | — | — | 0.66 | — | — | — | — | |
| SumEigvBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | — | — | — | — | — | 0.64 | — | — | — | |
| DegBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | — | — | — | — | — | — | 0.63 | — | — | |
| EccenBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | — | — | — | — | — | — | — | 0.65 | — | |
| SNNEBackbone=Llama2-13B, Access=Black-box2025.05 | 46.07 | — | — | — | — | — | — | — | — | — | — | — | — | 0.67 | |
| KLEfullBackbone=Phi-3-mini, Access=White-box2025.05 | 42.84 | 0.64 | — | — | — | — | — | — | — | — | — | — | — | — | |
| SEBackbone=Phi-3-mini, Access=White-box2025.05 | 42.84 | — | 0.63 | — | — | — | — | — | — | — | — | — | — | — | |
| NEBackbone=Phi-3-mini, Access=White-box2025.05 | 42.84 | — | — | 0.6 | — | — | — | — | — | — | — | — | — | — | |
| WSNNEBackbone=Phi-3-mini, Access=White-box2025.05 | 42.84 | — | — | — | 0.66 | — | — | — | — | — | — | — | — | — | |
| KLEheatBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | 0.63 | — | — | — | — | — | — | — | — | |
| DSEBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | — | 0.63 | — | — | — | — | — | — | — | |
| pTrueBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | — | — | 0.45 | — | — | — | — | — | — | |
| NumSetBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | — | — | — | 0.63 | — | — | — | — | — | |
| LexSimBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | — | — | — | — | 0.61 | — | — | — | — | |
| SumEigvBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | — | — | — | — | — | 0.62 | — | — | — | |
| DegBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | — | — | — | — | — | — | 0.63 | — | — | |
| EccenBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | — | — | — | — | — | — | — | 0.63 | — | |
| SNNEBackbone=Phi-3-mini, Access=Black-box2025.05 | 42.84 | — | — | — | — | — | — | — | — | — | — | — | — | 0.66 | |
| KLEfullBackbone=Llama2-7B, Access=White-box2025.05 | 41.73 | 0.63 | — | — | — | — | — | — | — | — | — | — | — | — | |
| SEBackbone=Llama2-7B, Access=White-box2025.05 | 41.73 | — | 0.61 | — | — | — | — | — | — | — | — | — | — | — | |
| NEBackbone=Llama2-7B, Access=White-box2025.05 | 41.73 | — | — | 0.59 | — | — | — | — | — | — | — | — | — | — | |
| WSNNEBackbone=Llama2-7B, Access=White-box2025.05 | 41.73 | — | — | — | 0.63 | — | — | — | — | — | — | — | — | — | |
| KLEheatBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | 0.62 | — | — | — | — | — | — | — | — | |
| DSEBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | — | 0.61 | — | — | — | — | — | — | — | |
| pTrueBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | — | — | 0.54 | — | — | — | — | — | — | |
| NumSetBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | — | — | — | 0.61 | — | — | — | — | — | |
| LexSimBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | — | — | — | — | 0.61 | — | — | — | — | |
| SumEigvBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | — | — | — | — | — | 0.62 | — | — | — | |
| DegBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | — | — | — | — | — | — | 0.61 | — | — | |
| EccenBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | — | — | — | — | — | — | — | 0.62 | — | |
| SNNEBackbone=Llama2-7B, Access=Black-box2025.05 | 41.73 | — | — | — | — | — | — | — | — | — | — | — | — | 0.63 | |
| KLEfullBackbone=gemma-2-2b, Access=White-box2025.05 | 39.28 | 0.64 | — | — | — | — | — | — | — | — | — | — | — | — | |
| SEBackbone=gemma-2-2b, Access=White-box2025.05 | 39.28 | — | 0.62 | — | — | — | — | — | — | — | — | — | — | — | |
| NEBackbone=gemma-2-2b, Access=White-box2025.05 | 39.28 | — | — | 0.59 | — | — | — | — | — | — | — | — | — | — | |
| WSNNEBackbone=gemma-2-2b, Access=White-box2025.05 | 39.28 | — | — | — | 0.63 | — | — | — | — | — | — | — | — | — | |
| KLEheatBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | 0.62 | — | — | — | — | — | — | — | — | |
| DSEBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | — | 0.63 | — | — | — | — | — | — | — | |
| pTrueBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | — | — | 0.54 | — | — | — | — | — | — | |
| NumSetBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | — | — | — | 0.62 | — | — | — | — | — | |
| LexSimBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | — | — | — | — | 0.6 | — | — | — | — | |
| SumEigvBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | — | — | — | — | — | 0.61 | — | — | — | |
| DegBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | — | — | — | — | — | — | 0.62 | — | — | |
| EccenBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | — | — | — | — | — | — | — | 0.62 | — | |
| SNNEBackbone=gemma-2-2b, Access=Black-box2025.05 | 39.28 | — | — | — | — | — | — | — | — | — | — | — | — | 0.63 |