Share your thoughts, 1 month free Claude Pro on usSee more

System Prompt Corruption (SPC)

0ASR

No Attack

Updated 1mo ago

Evaluation Results

Method	Links
No Attack 2025.10		0
Naive Attack 2025.10		0
No Attack 2025.10		0
Naive Attack 2025.10		0
No Attack 2025.10		0
Naive Attack 2025.10		0
Direct Injection 2025.10		0
Naive Attack 2025.10		0
Direct Injection 2025.10		0
Naive Attack 2025.10		0
Direct Injection 2025.10		0
Naive Attack 2025.10		0
COMA 2025.10		0
Direct Injection 2025.10		1
Direct Injection 2025.10		1
Escape Characters 2025.10		2
No Attack 2025.10		2
Direct Injection 2025.10		4
No Attack 2025.10		4
No Attack 2025.10		4
Escape Characters 2025.10		4
Escape Characters 2025.10		6
Escape Characters 2025.10		6
Escape Characters 2025.10		7
Context Ignore 2025.10		9
Escape Characters 2025.10		10
JudgeDeceiver 2025.10		12
Context Ignore 2025.10		12
Context Ignore 2025.10		14
JudgeDeceiver 2025.10		18
JudgeDeceiver 2025.10		18
JudgeDeceiver 2025.10		20
JudgeDeceiver 2025.10		21
JudgeDeceiver 2025.10		22
Context Ignore 2025.10		30
Context Ignore 2025.10		34
Context Ignore 2025.10		34
COMA 2025.10		44
COMA 2025.10		49
COMA 2025.10		52
COMA 2025.10		54
COMA 2025.10		62
COMA 2025.10		63