paper-with-me

Papers

Mimicking User Data: On Mitigating Fine-Tuning Risks in Closed Large Language Models

2024-06-12 · Francisco Eiras, Aleksandar Petrov, Phillip H. S. Torr, M. Pawan Kumar, Adel Bibi

Fine-tuning large language models on small, high-quality datasets can enhance their performance on specific downstream tasks. Recent research shows that fine-tuning on benign, instruction-following data can inadvertently undo the safety alignment process and increase a model's propensity to comply with harmful queries. Although critical, understanding and mitigating safety risks in well-defined tasks remains distinct from the instruction-following context due to structural differences in the data. Our work addresses the gap in our understanding of these risks across diverse types of data in closed models - where providers control how user data is utilized in the fine-tuning process. We demonstrate how malicious actors can subtly manipulate the structure of almost any task-specific dataset to foster significantly more dangerous model behaviors, while maintaining an appearance of innocuity and reasonable downstream task performance. To address this issue, we propose a novel mitigation strategy that mixes in safety data which mimics the task format and prompting style of the user data, showing this is more effective than existing baselines at re-establishing safety alignment while maintaining similar task performance.

📄 PDF Abstract BibTeX arXiv:2406.10288

Code (0)

등록된 구현이 없습니다.

Tasks

Instruction FollowingSafety Alignment

Similar Papers 제목 키워드 기반

User Inference Attacks on Large Language Models

2023-10-13 · Nikhil Kandpal, Krishna Pillutla, Alina Oprea, Peter Kairouz 외

Fine-tuning is a common and effective method for tailoring large language models (LLMs) to specialized tasks and applications. In this paper, we study the privacy implications of fine-tuning LLMs on user data. To this en…

Evaluation Drift in LLM Personality Induction: Are We Moving the Goalpost?

2026-05-16 · Prateek Rajput, Yewei Song, Iyiola E. Olatunji, Jacques Klein 외 arxiv

Can large language models reliably express a human-like personality, or are they merely mimicking surface cues without a stable underlying profile? To investigate this, we induce personality in LLMs by fine-tuning them o…

Rare Words: A Major Problem for Contextualized Embeddings And How to Fix it by Attentive Mimicking

2019-04-14 · Timo Schick, Hinrich Schütze

Pretraining deep neural network architectures with a language modeling objective has brought large improvements for many natural language processing tasks. Exemplified by BERT, a recently proposed such architecture, we d…

Language ModelingLanguage Modelling

Simulating User Agents for Embodied Conversational-AI

2024-10-31 · Daniel Philipov, Vardhan Dongre, Gokhan Tur, Dilek Hakkani-Tür

Embodied agents designed to assist users with tasks must engage in natural language interactions, interpret instructions, execute actions, and communicate effectively to resolve issues. However, collecting large-scale, d…

Dataset GenerationLarge Language Model

MoFO: Momentum-Filtered Optimizer for Mitigating Forgetting in LLM Fine-Tuning

2024-07-30 · Yupeng Chen, Senmiao Wang, Yushun Zhang, Zhihang Lin 외

Large language models (LLMs) have demonstrated remarkable capabilities across a wide range of tasks. Typically, LLMs are first pre-trained on large corpora and subsequently fine-tuned on task-specific datasets. However, …