List your company
ResearchFormerly WizWand

Machine Translation on English-French (Accuracy)

75.8AccuracyState vector (inn.)
-3.03217.43437.958.366Apr 17, 2024Jun 5, 2024Jul 25, 2024Sep 12, 2024Nov 1, 2024Dec 20, 2024Feb 8, 2025
Updated 8mo ago

Evaluation results

MethodLinks
State vector (inn.)Model=Llama-2-13B, Evaluation Protocol=Few-shot2024.0475.8
State vector (mom.)Model=Llama-2-13B, Evaluation Protocol=Few-shot2024.0475.6
+ LTVprompt=few-shot2025.0275.4
Function vectorModel=Llama-2-13B, Evaluation Protocol=Few-shot2024.0475.2
+ SVprompt=few-shot2025.0275
State vector (inn.)Model=Llama-2, Evaluation Protocol=Few-shot2024.0474.6
ICL baselineModel=Llama-2-13B, Evaluation Protocol=Few-shot2024.0474.5
+ ICVprompt=few-shot2025.0274.4
ICL baselineModel=Llama-2, Evaluation Protocol=Few-shot2024.0474.3
State vector (mom.)Model=Llama-2, Evaluation Protocol=Few-shot2024.0474.3
Task vectorModel=Llama-2, Evaluation Protocol=Few-shot2024.0473.8
Transformerprompt=few-shot2025.0271.9
State vector (inn.)Model=GPT-J, Evaluation Protocol=Few-shot2024.0470.9
Task vectorModel=GPT-J, Evaluation Protocol=Few-shot2024.0470.6
Task vectorModel=Llama-2-13B, Evaluation Protocol=Few-shot2024.0470.5
State vector (mom.)Model=GPT-J, Evaluation Protocol=Few-shot2024.0470.1
ICL baselineModel=GPT-J, Evaluation Protocol=Few-shot2024.0469.9
+ FVprompt=few-shot2025.0268
State vector (mom.)Model=Llama-2, Evaluation Protocol=Zero-shot2024.0467.5
State vector (inn.)Model=Llama-2, Evaluation Protocol=Zero-shot2024.0466.5
Function vectorModel=GPT-J, Evaluation Protocol=Few-shot2024.0465.8
Function vectorModel=Llama-2, Evaluation Protocol=Few-shot2024.0465.2
Task vectorModel=Llama-2, Evaluation Protocol=Zero-shot2024.0463.2
State vector (mom.)Model=Llama-2-13B, Evaluation Protocol=Zero-shot2024.0455.9
+ LTVprompt=zero-shot2025.0253.5
Standard Promptingprompt=few-shot2025.0252
State vector (inn.)Model=Llama-2-13B, Evaluation Protocol=Zero-shot2024.0450.5
Task vectorModel=Llama-2-13B, Evaluation Protocol=Zero-shot2024.0443.1
+ FVprompt=zero-shot2025.0240.2
State vector (mom.)Model=GPT-J, Evaluation Protocol=Zero-shot2024.0435.1
Task vectorModel=GPT-J, Evaluation Protocol=Zero-shot2024.0432.2
State vector (inn.)Model=GPT-J, Evaluation Protocol=Zero-shot2024.0431.7
Function vectorModel=GPT-J, Evaluation Protocol=Zero-shot2024.0429.1
+ SVprompt=zero-shot2025.0225.4
+ ICVprompt=zero-shot2025.0224.4
Transformerprompt=zero-shot2025.0223.4
Function vectorModel=Llama-2-13B, Evaluation Protocol=Zero-shot2024.0423.2
Function vectorModel=Llama-2, Evaluation Protocol=Zero-shot2024.0421.6
RegularModel=GPT-J, Evaluation Protocol=Zero-shot2024.047.2
RegularModel=Llama-2-13B, Evaluation Protocol=Zero-shot2024.040.2
RegularModel=Llama-2, Evaluation Protocol=Zero-shot2024.040.1
Standard Promptingprompt=zero-shot2025.020