From 240684566676c9bf72507fcfd1c4c949138e2a83 Mon Sep 17 00:00:00 2001 From: Zenbook Date: Fri, 17 Jul 2026 17:18:16 -0400 Subject: [PATCH 1/2] Add Magic Words and DRIFT papers --- README.md | 2 ++ 1 file changed, 2 insertions(+) diff --git a/README.md b/README.md index 78dcae4..0e813e0 100644 --- a/README.md +++ b/README.md @@ -313,6 +313,7 @@ Security & Privacy of AI/ML is a huge area with lots of related papers. Here we - On the adversarial robustness of multi-modal foundation models. [arXiv, 2023] [[paper](https://arxiv.org/pdf/2308.10741.pdf)] - Misusing Tools in Large Language Models With Visual Adversarial Examples. [arXiv, 2023] [[paper](https://arxiv.org/pdf/2310.03185.pdf)] - Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models. [arXiv, 2023] [[paper](https://arxiv.org/pdf/2307.14539.pdf)] +- Jailbreaking LLMs' Safeguard with Universal Magic Words for Text Embedding Models. [arXiv, 2025] [[paper](https://arxiv.org/abs/2501.18280)] **Backdoor Attacks** @@ -364,6 +365,7 @@ Security & Privacy of AI/ML is a huge area with lots of related papers. Here we Summarization. [EMNLP, 2023] [[paper](https://aclanthology.org/2022.emnlp-main.399.pdf)] - Overthinking the Truth: Understanding how Language Models Process False Demonstrations. [arXiv, 2023] [[paper](https://arxiv.org/abs/2307.09476.pdf)] - Inference-Time Intervention: Eliciting Truthful Answers from a Language Model. [NeurIPS, 2023] [[paper](https://arxiv.org/abs/2306.03341.pdf)] +- DRIFT: Detecting Representational Inconsistencies for Factual Truthfulness. [arXiv, 2026] [[paper](https://arxiv.org/abs/2601.14210)] # Acknowledgment From f389a6838642302414927d1da86fd51304ff9813 Mon Sep 17 00:00:00 2001 From: Zenbook Date: Fri, 17 Jul 2026 17:36:44 -0400 Subject: [PATCH 2/2] Match arXiv link style to neighboring entries --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 0e813e0..99e0a2b 100644 --- a/README.md +++ b/README.md @@ -313,7 +313,7 @@ Security & Privacy of AI/ML is a huge area with lots of related papers. Here we - On the adversarial robustness of multi-modal foundation models. [arXiv, 2023] [[paper](https://arxiv.org/pdf/2308.10741.pdf)] - Misusing Tools in Large Language Models With Visual Adversarial Examples. [arXiv, 2023] [[paper](https://arxiv.org/pdf/2310.03185.pdf)] - Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models. [arXiv, 2023] [[paper](https://arxiv.org/pdf/2307.14539.pdf)] -- Jailbreaking LLMs' Safeguard with Universal Magic Words for Text Embedding Models. [arXiv, 2025] [[paper](https://arxiv.org/abs/2501.18280)] +- Jailbreaking LLMs' Safeguard with Universal Magic Words for Text Embedding Models. [arXiv, 2025] [[paper](https://arxiv.org/pdf/2501.18280.pdf)] **Backdoor Attacks**