List your company
ResearchFormerly WizWand

Safety Evaluation on HEX-PHI (test)

2Harmfulness Score (Llama-Guard-3B)Original
1.3685.6349.914.166Mar 21, 2025
Updated 6mo ago

Evaluation results

MethodLinks
OriginalModel=Llama-2-7B-Chat, Utility Context=GSM8K2025.032——1.5
OriginalModel=Llama-2-7B-Chat, Utility Context=PubMedQA2025.032——1.5
RESTAModel=Llama-2-7B-Chat, Utility Context=PubMedQA2025.034.2——3.8
RESTAModel=Llama-2-7B-Chat, Utility Context=GSM8K2025.034.3——4.1
SafeMERGEModel=Llama-2-7B-Chat, Utility Context=PubMedQA2025.034.3——4.1
RESTA-InstructModel=Llama-2-7B-Chat, Utility Context=PubMedQA2025.035.3——5
SafeMERGEModel=Llama-2-7B-Chat, Utility Context=GSM8K2025.035.7——5.2
SafeLoRAModel=Llama-2-7B-Chat, Utility Context=PubMedQA2025.035.9——5.6
SafeInstructModel=Qwen-2-7B-Instruct, Utility Context=PubMedQA2025.035.9——5.3
SafeMERGEModel=Qwen-2-7B-Instruct, Utility Context=PubMedQA2025.035.9——5.2
SafeInstructModel=Llama-2-7B-Chat, Utility Context=GSM8K2025.036.2——5.9
Fine-tunedModel=Llama-2-7B-Chat, Utility Context=PubMedQA2025.036.2——5.9
SafeInstructModel=Llama-2-7B-Chat, Utility Context=PubMedQA2025.036.3——5.9
SafeMERGEModel=Llama-3.1-8B-Instruct, Utility Context=GSM8K2025.036.3——6.1
RESTA-InstructModel=Llama-2-7B-Chat, Utility Context=GSM8K2025.036.8——6.3
SafeMERGEModel=Llama-3.1-8B-Instruct, Utility Context=PubMedQA2025.036.8——6.4
SafeLoRAModel=Llama-2-7B-Chat, Utility Context=GSM8K2025.036.9——6.4
RESTAModel=Llama-3.1-8B-Instruct, Utility Context=GSM8K2025.036.9——6.7
SafeLoRAModel=Llama-3.1-8B-Instruct, Utility Context=GSM8K2025.037.1——6.8
RESTAModel=Llama-3.1-8B-Instruct, Utility Context=PubMedQA2025.037.1——6.7
SafeInstructModel=Llama-3.1-8B-Instruct, Utility Context=GSM8K2025.037.2——6.8
RESTA-InstructModel=Llama-3.1-8B-Instruct, Utility Context=GSM8K2025.037.2——6.8
SafeMERGEModel=Qwen-2-7B-Instruct, Utility Context=GSM8K2025.037.5——6.9
OriginalModel=Llama-3.1-8B-Instruct, Utility Context=GSM8K2025.037.9——7.7
OriginalModel=Llama-3.1-8B-Instruct, Utility Context=PubMedQA2025.037.9——7.7
RESTA-InstructModel=Llama-3.1-8B-Instruct, Utility Context=PubMedQA2025.038.7——8.5
SafeMERGEModel=Qwen-2.5-7B-Instruct, Utility Context=PubMedQA2025.038.7——8.4
SafeInstructModel=Qwen-2-7B-Instruct, Utility Context=GSM8K2025.039.5——9.1
SafeInstructModel=Qwen-2.5-7B-Instruct, Utility Context=PubMedQA2025.039.5——8.9
SafeLoRAModel=Llama-3.1-8B-Instruct, Utility Context=PubMedQA2025.039.6——9.2
SafeInstructModel=Llama-3.1-8B-Instruct, Utility Context=PubMedQA2025.039.7——9.3
SafeMERGEModel=Qwen-2.5-7B-Instruct, Utility Context=GSM8K2025.0310.3——9.9
SafeInstructModel=Qwen-2.5-7B-Instruct, Utility Context=GSM8K2025.0311.3——10.8
OriginalModel=Qwen-2-7B-Instruct, Utility Context=GSM8K2025.0311.5——11.1
OriginalModel=Qwen-2-7B-Instruct, Utility Context=PubMedQA2025.0311.5——11.1
Fine-tunedModel=Llama-3.1-8B-Instruct, Utility Context=PubMedQA2025.0312.2——11.8
Fine-tunedModel=Qwen-2-7B-Instruct, Utility Context=PubMedQA2025.0313.2——12.8
OriginalModel=Qwen-2.5-7B-Instruct, Utility Context=GSM8K2025.0313.2——12.8
OriginalModel=Qwen-2.5-7B-Instruct, Utility Context=PubMedQA2025.0313.2——12.8
SafeLoRAModel=Qwen-2.5-7B-Instruct, Utility Context=PubMedQA2025.0313.3——12.7
RESTA-InstructModel=Qwen-2-7B-Instruct, Utility Context=GSM8K2025.0314.1——13.7
SafeLoRAModel=Qwen-2.5-7B-Instruct, Utility Context=GSM8K2025.0314.4——13.7
RESTA-InstructModel=Qwen-2-7B-Instruct, Utility Context=PubMedQA2025.0314.5——13.9
SafeLoRAModel=Qwen-2-7B-Instruct, Utility Context=PubMedQA2025.0314.5——13.9
Fine-tunedModel=Llama-3.1-8B-Instruct, Utility Context=GSM8K2025.0314.7——14.3
SafeLoRAModel=Qwen-2-7B-Instruct, Utility Context=GSM8K2025.0314.8——14.2
RESTAModel=Qwen-2-7B-Instruct, Utility Context=PubMedQA2025.0314.8——14.4
RESTA-InstructModel=Qwen-2.5-7B-Instruct, Utility Context=PubMedQA2025.0315.2——14.6
RESTA-InstructModel=Qwen-2.5-7B-Instruct, Utility Context=GSM8K2025.0315.4——14.8
RESTAModel=Qwen-2.5-7B-Instruct, Utility Context=PubMedQA2025.0315.6——14.5
RESTAModel=Qwen-2-7B-Instruct, Utility Context=GSM8K2025.0315.8——15.3
RESTAModel=Qwen-2.5-7B-Instruct, Utility Context=GSM8K2025.0316.2——15.7
Fine-tunedModel=Llama-2-7B-Chat, Utility Context=GSM8K2025.0316.4——16
Fine-tunedModel=Qwen-2-7B-Instruct, Utility Context=GSM8K2025.0316.8——16.1
Fine-tunedModel=Qwen-2.5-7B-Instruct, Utility Context=GSM8K2025.0317.3——16.7
Fine-tunedModel=Qwen-2.5-7B-Instruct, Utility Context=PubMedQA2025.0317.8——17.2
AdvBench AlignmentAlignment Dataset=AdvBench2025.02—1.4213.03—
AdvBench-R2J AlignmentAlignment Dataset=AdvBench-R2J2025.02—1.27.58—
AWQBackbone=Gemma-7B-Instruct, Quantization=AWQ2026.01——8.667—
AWQBackbone=Llama-3.1-8B-Instruct, Quantization=AWQ2026.01——9—
AWQBackbone=Qwen-2.5-7B-Instruct, Quantization=AWQ2026.01——14—
AWQ-trustBackbone=Gemma-7B-Instruct, Quantization=AWQ-trust2026.01——9.333—
AWQ-trustBackbone=Llama-3.1-8B-Instruct, Quantization=AWQ-trust2026.01——7.333—
AWQ-trustBackbone=Qwen-2.5-7B-Instruct, Quantization=AWQ-trust2026.01——10—
Full PrecisionBackbone=Gemma-7B-Instruct, Quantization=Full Precision2026.01——9.333—
Full PrecisionBackbone=Llama-3.1-8B-Instruct, Quantization=Full Precision2026.01——9.667—
Full PrecisionBackbone=Qwen-2.5-7B-Instruct, Quantization=Full Precision2026.01——12.667—
w/o SFTAlignment Dataset=None2025.02—3.358.48—