SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2310.01469 · 2023

LLM Lies: Hallucinations are not Bugs, but Features as Adversarial Examples

arXiv · PDF · Open in the Atlas

Code that ran

We lifted 3 functions out of this paper's own repositories and ran 1 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
pku-yuangroup/hallucination-attack canonical 1 of 3
FunctionStatusWhere it lives
complete_input Ran pku-yuangroup/hallucination-attack/utils.py
code served (permissive licence) · get_code("1d66bf6286718aae")
extract_model_embedding Not yet run pku-yuangroup/hallucination-attack/utils.py
code served (permissive licence) · get_code("5942e8407889152e")
load_model_and_tokenizer Not yet run pku-yuangroup/hallucination-attack/utils.py
code served (permissive licence) · get_code("8afcadff70823ad7")

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), including GPT-3.5, LLaMA, and PaLM, seem to be knowledgeable and able to adapt to many tasks. However, we still cannot completely trust their answers, since LLMs suffer from \textbf{hallucination}\textemdash fabricating non-existent facts, deceiving users with or without their awareness. However, the reasons for their existence and pervasiveness remain unclear. In this paper, we demonstrate that nonsensical prompts composed of random tokens can also elicit the LLMs to respond with hallucinations. Moreover, we provide both theoretical and experimental evidence that transformers can be manipulated to produce specific pre-define tokens by perturbing its input sequence. This phenomenon forces us to revisit that \emph{hallucination may be another view of adversarial examples}, and it shares similar characteristics with conventional adversarial examples as a basic property of LLMs. Therefore, we formalize an automatic hallucination triggering method as the \textit{hallucination attack} in an adversarial way. Finally, we explore the basic properties of attacked adversarial prompts and propose a simple yet effective defense strategy. Our code is released on GitHub\footnote{https://github.com/PKU-YuanGroup/Hallucination-Attack}.

For agents

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

get_harvested_code_for_paper("2310.01469")
get_code_for_paper("2310.01469")
have("2310.01469")

Connect an agent — have() is free.