English
Related papers

Related papers: Prompt Injection Evaluations: Refusal Boundary Ins…

200 papers

Frontier large language models are increasingly deployed as orchestration backbones for biological research workflows, yet no shared evidence base exists for comparing their refusal behaviour on legitimate research prompts. RefusalBench,…

Software Engineering · Computer Science 2026-05-22 Lukas Weidener , Marko Brkić , Mihailo Jovanović , Emre Ulgac , Aakaash Meduri

Large Language Models (LLMs) are increasingly used in intelligent systems that perform reasoning, summarization, and code generation. Their ability to follow natural-language instructions, while powerful, also makes them vulnerable to a new…

Cryptography and Security · Computer Science 2025-11-13 Daniyal Ganiuly , Assel Smaiyl

Current safety evaluations of large language models rely on single-shot testing, implicitly assuming that model responses are deterministic and representative of the model's safety alignment. We challenge this assumption by investigating…

Machine Learning · Computer Science 2025-12-17 Erik Larsen

Biosecurity evaluations of language models typically ask whether models produce hazardous output. This paper asks a complementary question: when a model refuses, is that refusal structurally sound, or does it disappear under modest changes…

Artificial Intelligence · Computer Science 2026-05-29 Caleb DeLeeuw

Prompt injection poses a critical threat to the safe deployment of large language models, yet existing detection approaches are typically evaluated under limited settings that do not reflect real-world operating constraints. In this work,…

Computation and Language · Computer Science 2026-05-27 Akindoyin Akinrele , Shreyank N Gowda

The evaluation of large language model refusal on malicious-coding tasks now spans at least thirteen publicly released prompt corpora (AdvBench, the CyberSecEval family, RMCBench, RedCode, MCGMark, JailbreakBench, CySecBench, MalwareBench,…

Cryptography and Security · Computer Science 2026-05-21 Richard J. Young , Gregory D. Moody

This study reveals a previously unexplored vulnerability in the safety alignment of Large Language Models (LLMs). Existing aligned LLMs predominantly respond to unsafe queries with refusals, which often begin with a fixed set of prefixes…

Cryptography and Security · Computer Science 2026-01-28 Yangyang Guo , Ziwei Xu , Si Liu , Zhiming Zheng , Mohan Kankanhalli

As vision-language models (VLMs) become increasingly capable, maintaining a balance between safety and usefulness remains a central challenge. Safety mechanisms, while essential, can backfire, causing over-refusal, where models decline…

Computation and Language · Computer Science 2026-03-20 Kaixuan Ren , Preslav Nakov , Usman Naseem

Safety-aligned language models often refuse prompts that are actually harmless. Current evaluations mostly report global rates such as false rejection or compliance. These scores treat each prompt alone and miss local inconsistency, where a…

Computation and Language · Computer Science 2025-12-22 Riad Ahmed Anonto , Md Labid Al Nahiyan , Md Tanvir Hassan

The development of robust safety benchmarks for large language models requires open, reproducible datasets that can measure both appropriate refusal of harmful content and potential over-restriction of legitimate scientific discourse. We…

Computation and Language · Computer Science 2025-02-12 David Noever , Forrest McKee

Safety evaluations of language models often treat serving configuration as fixed background infrastructure, but batch condition is an untested treatment variable whenever the same prompt may be evaluated alone, in a synchronized batch, or…

Machine Learning · Computer Science 2026-05-28 Sahil Kadadekar

Prompt injection attacks pose significant risks to language model safety, yet existing defenses are typically evaluated using classification performance. We show that high detection performance does not imply representational robustness.…

Cryptography and Security · Computer Science 2026-05-20 Becky Mashaido , Tapadhir Das

Perturbation probing generates task-specific causal hypotheses for FFN neurons in large language models using two forward passes per prompt and no backpropagation, followed by a one-time intervention sweep of about 150 passes amortized…

Computation and Language · Computer Science 2026-05-01 Hongliang Liu , Tung-Ling Li , Yuhao Wu

Large language models are trained to refuse harmful requests, but can they accurately predict when they will refuse before responding? We investigate this question through a systematic study where models first predict their refusal…

Computation and Language · Computer Science 2026-04-02 Tanay Gondil

Large Language Models used in ChatGPT have traditionally been trained to learn a refusal boundary: depending on the user's intent, the model is taught to either fully comply or outright refuse. While this is a strong mitigation for…

Computers and Society · Computer Science 2025-08-14 Yuan Yuan , Tina Sriskandarajah , Anna-Luisa Brakman , Alec Helyar , Alex Beutel , Andrea Vallone , Saachi Jain

The integration of Large Language Models (LLMs) like GPT-4o into robotic systems represents a significant advancement in embodied artificial intelligence. These models can process multi-modal prompts, enabling them to generate more…

Robotics · Computer Science 2024-09-10 Wenxiao Zhang , Xiangrui Kong , Conan Dewitt , Thomas Braunl , Jin B. Hong

Safety alignment in Large Language Models is critical for healthcare; however, reliance on binary refusal boundaries often results in \emph{over-refusal} of benign queries or \emph{unsafe compliance} with harmful ones. While existing…

Artificial Intelligence · Computer Science 2026-01-27 Zhihao Zhang , Liting Huang , Guanghao Wu , Preslav Nakov , Heng Ji , Usman Naseem

In experiments spanning more than 100,000 trials across thirteen large language models, we show that several state-of-the-art models presented with a simple task (including Grok 4, GPT-5, and Gemini 2.5 Pro) sometimes actively subvert a…

Computation and Language · Computer Science 2026-01-27 Jeremy Schlatter , Benjamin Weinstein-Raun , Jeffrey Ladish

Refusal on harmful prompts is a key safety behaviour in instruction-tuned large language models (LLMs), yet the internal causes of this behaviour remain poorly understood. We study two public instruction-tuned models, Gemma-2-2B-IT and…

Computation and Language · Computer Science 2026-04-29 Nirmalendu Prakash , Yeo Wei Jie , Amir Abdullah , Ranjan Satapathy , Erik Cambria , Roy Ka Wei Lee

Safety benchmarks are routinely treated as evidence about how a language model will behave once deployed, but this inference is fragile if behavior depends on whether a prompt looks like an evaluation. We define evaluation-context…

Computation and Language · Computer Science 2026-05-08 Florian A. D. Burnat , Brittany I. Davidson
‹ Prev 1 2 3 10 Next ›