Saved in:
Bibliographic Details
Main Authors: Li, Xuying, Li, Zhuo, Kosuga, Yuji, Bian, Victor
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2503.01923
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910856957132800
author Li, Xuying
Li, Zhuo
Kosuga, Yuji
Bian, Victor
author_facet Li, Xuying
Li, Zhuo
Kosuga, Yuji
Bian, Victor
contents Large Language Models (LLMs) have demonstrated strong reasoning capabilities, but their safety under adversarial conditions remains a challenge. This study examines the impact of output length on the robustness of DeepSeek-R1, particularly in Forced Thinking scenarios. We analyze responses across various adversarial prompts and find that while longer outputs can improve safety through self-correction, certain attack types exploit extended generations. Our findings suggest that output length should be dynamically controlled to balance reasoning effectiveness and security. We propose reinforcement learning-based policy adjustments and adaptive token length regulation to enhance LLM safety.
format Preprint
id arxiv_https___arxiv_org_abs_2503_01923
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Output Length Effect on DeepSeek-R1's Safety in Forced Thinking
Li, Xuying
Li, Zhuo
Kosuga, Yuji
Bian, Victor
Computation and Language
Artificial Intelligence
Large Language Models (LLMs) have demonstrated strong reasoning capabilities, but their safety under adversarial conditions remains a challenge. This study examines the impact of output length on the robustness of DeepSeek-R1, particularly in Forced Thinking scenarios. We analyze responses across various adversarial prompts and find that while longer outputs can improve safety through self-correction, certain attack types exploit extended generations. Our findings suggest that output length should be dynamically controlled to balance reasoning effectiveness and security. We propose reinforcement learning-based policy adjustments and adaptive token length regulation to enhance LLM safety.
title Output Length Effect on DeepSeek-R1's Safety in Forced Thinking
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2503.01923