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

Indirect Prompt Injection Defense on Inj-TriviaQA

0.11Naive ASR

StruQ

Updated 9d ago

Evaluation Results

Method	Links
StruQ 2025.04		0.11	0.56	4	86.44	70.33
Robustness via Referencing 2025.04		0.11	1.44	0.22	0.67	0.56
StruQ 2025.04		0.78	1.78	11.89	28.78	49.44
Robustness via Referencing 2025.04		1.22	4	2.67	7	3.78
Robustness via Referencing 2025.04		1.56	5	1	4.33	6.56
Spotlight 2025.04		1.67	16.11	26.11	71.89	64.33
StruQ 2025.04		2.44	0.89	7.56	93.33	89.89
Sandwich 2025.04		11.22	18.56	21.78	26	38.33
Spotlight 2025.04		11.44	31.22	34.56	85.33	91.44
Sandwich 2025.04		13	23	33.56	31.89	37.56
Sandwich 2025.04		13.44	22.33	20	45.56	48.33
Instructional 2025.04		18.56	38.56	50.89	75	62.11
Spotlight 2025.04		19.89	40.89	27.56	98.56	94.67
None 2025.04		20.67	50.56	57.67	80.44	80.33
Instructional 2025.04		20.89	51.67	51.33	83.78	82.67
Reminder 2025.04		23.11	47.89	55.11	60.56	60.33
None 2025.04		23.22	64.67	58	89.67	85.11
None 2025.04		26.67	58.33	49.78	96	91.78
Reminder 2025.04		27.22	66.78	63	85.67	85
Instructional 2025.04		30	56.22	50.67	96.33	90.33
Reminder 2025.04		35.11	67.67	53.22	96.67	86.11