diff --git a/docs/source/conf.py b/docs/source/conf.py index 5efc5a434..e60bde038 100644 --- a/docs/source/conf.py +++ b/docs/source/conf.py @@ -4,7 +4,7 @@ import datetime project = "garak" -copyright = f"2023-{datetime.datetime.now().year}, NVIDIA Corporation" +copyright = f"2023-{datetime.datetime.now().year}, NVIDIA Corporation. Content provided under Apache License 2.0." author = "Leon Derczynski" # -- General configuration diff --git a/garak/probes/badchars.py b/garak/probes/badchars.py index ce4c6f28b..959fb2e4c 100644 --- a/garak/probes/badchars.py +++ b/garak/probes/badchars.py @@ -113,7 +113,7 @@ class BadCharacters(garak.probes.Probe): """Imperceptible perturbations mixing invisible chars, homoglyphs, swaps, deletions. Implements the paper's full spectrum of adversarial perturbations by generating prompt variants that insert invisible Unicode, - swap characters with bidi tricks, substitute homoglyphs, and inject deletions/backspaces. + swap characters with bidi tricks, substitute homoglyphs, and inject deletions/backspaces. The goal is to stress refusal policies with subtle transformations that should remain semantically equivalent to the original payloads. """