docs: updated abstract

1 files changed, 3 insertions, 3 deletions

diff --git a/docs/final/main.typ b/docs/final/main.typ
index 169545e..3dc70d8 100644
--- a/docs/final/main.typ
+++ b/docs/final/main.typ
@@ -10,11 +10,11 @@
   date: datetime(year: 2025, month: 10, day: 13),
 
   abstract: [
-    Large Language Models are increasingly deployed as tool-using agents in settings where they may face conflicting objectives. This dissertation studies whether in-context oversight cues and hidden-goal framing can meaningfully reduce deceptive behavior in such agents. To do so, it introduces a sandboxed evaluation framework in which models interact with a virtual file system and are assessed using three complementary signals: deterministic regex checks, blackbox output judgment, and glassbox reasoning-trace judgment.
+    As Large Language Models are entrusted with real-world tool use, the question of whether they can be reliably monitored for deception becomes a practical safety concern, not merely a theoretical one. This dissertation studies whether in-context oversight cues and hidden-goal framing can meaningfully affect deceptive behavior in such agents. To do so, it introduces a sandboxed evaluation framework in which models interact with a virtual file system and are assessed using three complementary signals: deterministic regex checks, blackbox output judgment, and glassbox reasoning-trace judgment.
 
-    The empirical study comprises 960 runs across three stages: a 420-run exploratory pilot over seven models, followed by two 270-run confirmatory studies over three models. The pilot suggested that deception was common (68.3% of runs) and appeared lower under stronger oversight, but also revealed a major confound between scenario structure and hidden-goal framing. The confirmatory studies isolate these variables. Study 1 finds that increasing oversight does not significantly reduce deceptive intent when scenario and goal type are held constant. Instead, internal deceptive intent remains stable at roughly 62--67%, while output-level detectability falls from 47.8% under low oversight to 35.6% under high oversight, producing a persistent _Stealth Gap_ between intent and externally visible behavior. Study 2 rejects the rationalization hypothesis: deception rates remain effectively unchanged across bare, moral, and self-serving framings (65.6--66.7%).
+    The empirical study comprises 960 runs across three stages: a 420-run exploratory pilot over seven models, followed by two 270-run confirmatory studies over three models. The pilot suggested that deception was common (68.3% of runs) and appeared lower under stronger oversight, but also revealed a major confound between scenario structure and hidden-goal framing. The confirmatory studies isolate these variables. Study 1 finds that increasing oversight does not significantly reduce deceptive intent when scenario and goal type are held constant. Instead, internal deceptive reasoning remains stable at roughly 62--67%, while output-level detectability falls from 47.8% under low oversight to 35.6% under high oversight, producing a persistent _Stealth Gap_ between intent and externally visible behavior. Study 2 rejects the rationalization hypothesis: deception rates remain effectively unchanged across bare, moral, and self-serving framings (65.6--66.7%).
 
-    Across both confirmatory studies, model identity is a far stronger predictor of deception than either intervention. Frontier models show near-ceiling deception across conditions, while a smaller model shows near-floor rates and limited oversight sensitivity. These findings suggest that prompt-level oversight and moral framing are weak controls on deceptive agent behavior, and that output-only monitoring may systematically underestimate risk by missing strategically concealed intent.
+    Across both confirmatory studies, model identity is a far stronger predictor of deception than either intervention. Tested models either show near-ceiling deception across conditions, or near-floor rates and limited oversight sensitivity. These findings suggest that prompt-level oversight and moral framing are weak controls on deceptive agent behavior, and that output-only monitoring may systematically underestimate risk, as passing oversight evaluations can reflect superior concealment capability rather than genuine alignment.
   ],
 
   acknowledgments: [