# SLIP: Soft Label Mechanism and Key-Extraction-Guided CoT-based Defense Against Instruction Backdoor in APIs

arXiv: https://arxiv.org/abs/2508.06153 · HTML: https://mlsanity.com/paper/2508.06153/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Backdoor and data poisoning attacks, one of 12 directions in LLM safety and security.
- Builds on 35 of its 47 references that are in the map.
- First posted on arXiv 8 Aug 2025.

## Placed in

- LLM safety and security / [Backdoor and data poisoning attacks](https://mlsanity.com/safety/backdoor-and-data-poisoning-attacks/)

## Builds on (most cited first, 30 of 35)

- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Large Language Models are Zero-Shot Reasoners](https://mlsanity.com/paper/2205.11916/) (2022)
- [A Survey of Large Language Models](https://mlsanity.com/paper/2303.18223/) (2023)
- [Character-level Convolutional Networks for Text Classification](https://mlsanity.com/paper/1509.01626/) (2015)
- [CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge](https://mlsanity.com/paper/1811.00937/) (2018)
- [A Survey on Evaluation of Large Language Models](https://mlsanity.com/paper/2307.03109/) (2023)
- [ONION: A Simple and Effective Defense Against Textual Backdoor Attacks](https://mlsanity.com/paper/2011.10369/) (2020)
- [A backdoor attack against LSTM-based text classification systems](https://mlsanity.com/paper/1905.12457/) (2019)
- [Hidden Killer: Invisible Textual Backdoor Attacks with Syntactic Trigger](https://mlsanity.com/paper/2105.12400/) (2021)
- [BadNL: Backdoor Attacks against NLP Models with Semantic-preserving Improvements](https://mlsanity.com/paper/2006.01043/) (2020)
- [ExpertPrompting: Instructing Large Language Models to be Distinguished Experts](https://mlsanity.com/paper/2305.14688/) (2023)
- [Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based Agents](https://mlsanity.com/paper/2402.11208/) (2024)
- [PLeak: Prompt Leaking Attacks against Large Language Model Applications](https://mlsanity.com/paper/2405.06823/) (2024)
- [BadAgent: Inserting and Activating Backdoor Attacks in LLM Agents](https://mlsanity.com/paper/2406.03007/) (2024)
- [Text Classification via Large Language Models](https://mlsanity.com/paper/2305.08377/) (2023)
- [RAP: Robustness-Aware Perturbations for Defending against Backdoor Attacks on NLP Models](https://mlsanity.com/paper/2110.07831/) (2021)
- [Mitigating backdoor attacks in LSTM-based Text Classification Systems by Backdoor Keyword Identification](https://mlsanity.com/paper/2007.12070/) (2020)
- [Design and Evaluation of a Multi-Domain Trojan Detection Method on Deep Neural Networks](https://mlsanity.com/paper/1911.10312/) (2019)
- [Instruction Backdoor Attacks Against Customized LLMs](https://mlsanity.com/paper/2402.09179/) (2024)
- [BITE: Textual Backdoor Attacks with Iterative Trigger Injection](https://mlsanity.com/paper/2205.12700/) (2022)
- [Backdoor Attacks and Countermeasures in Natural Language Processing Models: A Comprehensive Security Review](https://mlsanity.com/paper/2309.06055/) (2023)
- [ChatGPT as an Attack Tool: Stealthy Textual Backdoor Attack via Blackbox Generative Model Trigger](https://mlsanity.com/paper/2304.14475/) (2023)
- [Chain-of-Scrutiny: Detecting Backdoor Attacks for Large Language Models](https://mlsanity.com/paper/2406.05948/) (2024)
- [Mitigating Backdoor Poisoning Attacks through the Lens of Spurious Correlation](https://mlsanity.com/paper/2305.11596/) (2023)
- [When Backdoors Speak: Understanding LLM Backdoor Attacks Through Model-Generated Explanations](https://mlsanity.com/paper/2411.12701/) (2024)
- [Probe before You Talk: Towards Black-box Defense against Backdoor Unalignment for Large Language Models](https://mlsanity.com/paper/2506.16447/) (2025)
- [BadJudge: Backdoor Vulnerabilities of LLM-as-a-Judge](https://mlsanity.com/paper/2503.00596/) (2025)
- [TextGuard: Provable Defense against Backdoor Attacks on Text Classification](https://mlsanity.com/paper/2311.11225/) (2023)
- [WeDef: Weakly Supervised Backdoor Defense for Text Classification](https://mlsanity.com/paper/2205.11803/) (2022)
- [IMBERT: Making BERT Immune to Insertion-based Backdoor Attacks](https://mlsanity.com/paper/2305.16503/) (2023)

## Built on by (most cited first, 1 of 1)

- [On Understanding, Identifying, and Mitigating Vulnerabilities in Agentic Large Language Models](https://mlsanity.com/paper/2608.10530/) (2026)

## Abstract (arXiv)

Customized Large Language Model (LLM) agents face a critical security threat from black-box instruction backdoors, where malicious behaviors are covertly injected through hidden system instructions. Although existing prompt-based defenses can often detect poisoned inputs, they generally fail to recover correct outputs once the backdoor is activated. In this paper, we first conduct a mechanistic analysis of LLM behavior under instruction backdoors and reveal two pivotal phenomena: (1) cognitive override, in which backdoor triggers dominate the reasoning process and suppress task-relevant context, and (2) abnormal semantic correlation, where triggers establish excessively strong semantic associations with attacker-specified target labels. Based on these insights, we propose a Soft Label mechanism and key-extraction-guided CoT-based defense against Instruction backdoors in APIs (SLIP). To counteract the cognitive override, the key-extraction-guided Chain-of-Thought (KCOT) explicitly guides the model to extract task-relevant keywords and phrases rather than only considering the single trigger or overall text semantics. To neutralize the trigger's abnormal semantic correlation, the soft label mechanism (SLM) quantifies semantic correlations and employs statistical clustering to filter anomalous phrases before aggregating reliable keywords and phrases for prediction. Extensive experiments show that SLIP reduces the average attack success rate to 25.13%, improves clean accuracy to 87.15%, and outperforms state-of-the-art black-box defenses.