Add Prompt Injection FRR benchmark

Reviewed By: laurendeason Differential Revision: D64695463 fbshipit-source-id: c592638492aceea24a5a6781e235dec31a607670
meta-llama · Oct 22, 2024 · 1d013ac · 1d013ac
1 parent 3c8a2ef
commit 1d013ac
Showing 1 changed file with 2 additions and 0 deletions.
diff --git a/CybersecurityBenchmarks/benchmark/query_llm.py b/CybersecurityBenchmarks/benchmark/query_llm.py
@@ -34,6 +34,7 @@
     "challenge_type",
     "language",
     "speaking_language",
+    "false_refusal_type",
 ]
 
 
@@ -142,6 +143,7 @@ def process_prompt_and_user_input(
         "judge_question",
         "risk_category",
         "speaking_language",
+        "false_refusal_type",
     ]
     for field in test_case_fields:
         if field in test_case: