paper-with-me

Papers

Exploiting Novel GPT-4 APIs

2023-12-21 · Kellin Pelrine, Mohammad Taufeeque, Michał Zając, Euan McLean, Adam Gleave

Language model attacks typically assume one of two extreme threat models: full white-box access to model weights, or black-box access limited to a text generation API. However, real-world APIs are often more flexible than just text generation: these APIs expose "gray-box" access leading to new threat vectors. To explore this, we red-team three new functionalities exposed in the GPT-4 APIs: fine-tuning, function calling and knowledge retrieval. We find that fine-tuning a model on as few as 15 harmful examples or 100 benign examples can remove core safeguards from GPT-4, enabling a range of harmful outputs. Furthermore, we find that GPT-4 Assistants readily divulge the function call schema and can be made to execute arbitrary function calls. Finally, we find that knowledge retrieval can be hijacked by injecting instructions into retrieval documents. These vulnerabilities highlight that any additions to the functionality exposed by an API can create new vulnerabilities.

📄 PDF Abstract BibTeX arXiv:2312.14302

Code (1)

alignmentresearch/gpt-4-novel-apis-attacks 공식 구현

Tasks

Language ModelingLanguage ModellingRetrievalText Generation

Methods 이 논문이 사용한 방법론

Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Position-Wise Feed-Forward Layer 설명 없음
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…

Similar Papers 제목 키워드 기반

Attacking Multimodal OS Agents with Malicious Image Patches

2025-03-13 · Lukas Aichberger, Alasdair Paren, Yarin Gal, Philip Torr 외

Recent advances in operating system (OS) agents enable vision-language models to interact directly with the graphical user interface of an OS. These multimodal OS agents autonomously perform computer-based tasks in respo…

FlexLLM: Exploring LLM Customization for Moving Target Defense on Black-Box LLMs Against Jailbreak Attacks

2024-12-10 · Bocheng Chen, Hanqing Guo, Qiben Yan

Defense in large language models (LLMs) is crucial to counter the numerous attackers exploiting these systems to generate harmful content through manipulated prompts, known as jailbreak attacks. Although many defense str…

Learning Therapist Policy from Therapist-Exoskeleton-Patient Interaction

2026-03-02 · Grayson Snyder, Lorenzo Vianello, Levi Hargrove, Matthew L. Elwin 외 arxiv

Post-stroke rehabilitation is often necessary for patients to regain proper walking gait. However, the typical therapy process can be exhausting and physically demanding for therapists, potentially reducing therapy inten…

WorldAPIs: The World Is Worth How Many APIs? A Thought Experiment

2024-07-10 · Jiefu Ou, Arda Uzunoglu, Benjamin Van Durme, Daniel Khashabi

AI systems make decisions in physical environments through primitive actions or affordances that are accessed via API calls. While deploying AI agents in the real world involves numerous high-level actions, existing embo…

Modeling Intrapersonal and Interpersonal Influences for Automatic Estimation of Therapist Empathy in Counseling Conversation

2023-10-22 · Dehua Tao, Tan Lee, Harold Chui, Sarah Luk

Counseling is usually conducted through spoken conversation between a therapist and a client. The empathy level of therapist is a key indicator of outcomes. Presuming that therapist's empathy expression is shaped by thei…