Pith. sign in

REVIEW 3 cited by

ICLShield: Exploring and Mitigating In-Context Learning Backdoor Attacks

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2507.01321 v1 pith:LAELDUND submitted 2025-07-02 cs.LG cs.AIcs.CR

ICLShield: Exploring and Mitigating In-Context Learning Backdoor Attacks

classification cs.LG cs.AIcs.CR
keywords backdoorllmsattacksdemonstrationsmethodadaptabilityconceptconcepts
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

In-context learning (ICL) has demonstrated remarkable success in large language models (LLMs) due to its adaptability and parameter-free nature. However, it also introduces a critical vulnerability to backdoor attacks, where adversaries can manipulate LLM behaviors by simply poisoning a few ICL demonstrations. In this paper, we propose, for the first time, the dual-learning hypothesis, which posits that LLMs simultaneously learn both the task-relevant latent concepts and backdoor latent concepts within poisoned demonstrations, jointly influencing the probability of model outputs. Through theoretical analysis, we derive an upper bound for ICL backdoor effects, revealing that the vulnerability is dominated by the concept preference ratio between the task and the backdoor. Motivated by these findings, we propose ICLShield, a defense mechanism that dynamically adjusts the concept preference ratio. Our method encourages LLMs to select clean demonstrations during the ICL phase by leveraging confidence and similarity scores, effectively mitigating susceptibility to backdoor attacks. Extensive experiments across multiple LLMs and tasks demonstrate that our method achieves state-of-the-art defense effectiveness, significantly outperforming existing approaches (+26.02% on average). Furthermore, our method exhibits exceptional adaptability and defensive performance even for closed-source models (e.g., GPT-4).

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 3 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. TimeGuard: Channel-wise Pool Training for Backdoor Defense in Time Series Forecasting

    cs.CR 2026-05 unverdicted novelty 7.0

    TimeGuard defends time series forecasting against backdoors via channel-wise pool training initialized by time-aware criteria and expanded with distance-regularized loss selection, improving poisoned MAE by 1.96x whil...

  2. TimeGuard: Channel-wise Pool Training for Backdoor Defense in Time Series Forecasting

    cs.CR 2026-05 unverdicted novelty 6.0

    TimeGuard employs channel-wise pool training initialized with time-aware criteria and distance-regularized loss selection to defend time series forecasting against backdoor attacks, improving robustness by 1.96x while...

  3. A Patch-based Cross-view Regularized Framework for Backdoor Defense in Multimodal Large Language Models

    cs.CV 2026-04 unverdicted novelty 5.0

    A patch-augmented cross-view regularization method reduces backdoor attack success rates in multimodal LLMs by enforcing output differences between original and perturbed views while using entropy constraints to prese...