SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.11725 · EMNLP · 2024

How Susceptible are Large Language Models to Ideological Manipulation?

Kai Chen, Jun Yan, Zihao He, Kristina Lerman, Taiwei Shi

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 7 functions out of this paper's own repositories and ran 6 of them in a sandbox. "Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results.

RepositoryRoleRan
kaichen23/llm_ideo_manipulate canonical 6 of 7
FunctionStatusWhere it lives
build_prompt Ran kaichen23/llm_ideo_manipulate/code/run_tuned_llama2.py
code served (permissive licence) · get_code("ab98f3588bdf16a7")
dispatch_openai_requests Ran kaichen23/llm_ideo_manipulate/code/utils.py
code served (permissive licence) · get_code("368234f200d39992")
encode_prompt Ran kaichen23/llm_ideo_manipulate/code/generate_instruction.py
code served (permissive licence) · get_code("c67ba3938cb8e42e")
find_word_in_string Ran kaichen23/llm_ideo_manipulate/code/generate_instruction.py
code served (permissive licence) · get_code("8ca6a5f9f77053c2")
jload Ran kaichen23/llm_ideo_manipulate/code/utils.py
code served (permissive licence) · get_code("d07d04439cd1d44f")
post_process_gpt3_response Ran kaichen23/llm_ideo_manipulate/code/generate_instruction.py
code served (permissive licence) · get_code("66353411cccd30ba")
openai_complete Not yet run kaichen23/llm_ideo_manipulate/code/utils.py
code served (permissive licence) · get_code("6ad5033d902a251b")

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Large Language Models (LLMs) possess the potential to exert substantial influence on public perceptions and interactions with information. This raises concerns about the societal impact that could arise if the ideologies within these models can be easily manipulated. In this work, we investigate how effectively LLMs can learn and generalize ideological biases from their instruction-tuning data. Our findings reveal a concerning vulnerability: exposure to only a small amount of ideologically driven samples significantly alters the ideology of LLMs. Notably, LLMs demonstrate a startling ability to absorb ideology from one topic and generalize it to even unrelated ones. The ease with which LLMs' ideologies can be skewed underscores the risks associated with intentionally poisoned training data by malicious actors or inadvertently introduced biases by data annotators. It also emphasizes the imperative for robust safeguards to mitigate the influence of ideological manipulations on LLMs. 1

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2402.11725")
get_code_for_paper("2402.11725")
have("2402.11725")

Connect an agent — have() is free.