arXiv AI September 18, 2026 By Wenyuan Wu, Yuan Sun, Yingke Chen, Chao Su, Xi Peng, Dezhong Peng, Xu Wang

Information-Geometric Inverse Distillation for Enhancing Adversarial Transferability

The paper introduces Inverse Knowledge Distillation (IKD), an attack‑agnostic technique that enhances adversarial transferability by maximizing the discrepancy between benign and adversarial prediction distributions on a surrogate model. IKD employs a CE/KL‑equivalent soft‑label objective to push adversarial predictions away from a fixed benign anchor, leveraging Fisher‑sensitive surrogate directions. The authors provide theoretical analysis showing CE and KL induce identical gradients, derive a lower bound on Fisher‑subspace overlap, and demonstrate through extensive ImageNet experiments that IKD consistently improves black‑box attack performance across CNN, ViT, and defended models.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv AI

Jun 10

Improving Adversarial Transferability on Vision-Language Pre-training Models via Surrogate-Specific Bias Correction

arXiv:2606. 10571v1 Announce Type: cross Abstract: Adversarial examples reveal vulnerabilities in Vision-Language Pre-training (VLP) models and provide insights for improving robustness.

By Lijia Yu, Jiuxin Cao, Yuchen Qiang, Changhao Chen, Yifei Huang, Bo Liu

llms multimodal safety

arXiv Machine Learning

Aug 4

DeepDefense: Robust Learning via Layer-Wise Gradient-Feature Alignment

arXiv:2511. 13749v2 Announce Type: replace Abstract: Deep neural networks are known to be vulnerable to adversarial perturbations, which are small, carefully crafted inputs that lead to incorrect predictions.

By Ci Lin, Tet Yeap, Iluju Kiringa

safety

arXiv Machine Learning

Aug 21

Merge Now, Regret Later: The Hidden Cost of Model Merging Is Adversarial Transferability

arXiv:2509. 23689v2 Announce Type: replace Abstract: Model Merging (MM) has proven to be an effective alternative to multi-task learning, where several fine-tuned models are merged, without access to the tasks' training data, into one model that retains performance across different tasks.

By Mauro Conti, Ankit Gangwal, Aaryan Ajay Sharma

fine-tuning safety

Hugging Face Trending Papers

Jun 2

Investigating Adversarial Robustness of Multi-modal Large Language Models

Multi-modal Large Language Models (MLLMs) achieve strong performance on vision-language tasks, but incorporating visual inputs through a vision encoder (e. g.

llms rag fine-tuning multimodal safety

arXiv Machine Learning

Jun 11

Latent Geometric Chords for Query-Efficient Decision-Based Adversarial Attacks

arXiv:2605. 31219v2 Announce Type: replace-cross Abstract: While decision-based black-box adversarial attacks present a severe security threat, current methodologies suffer from fundamental limitations.

By Ei Hmue Khine, Yao Li, Jiebao Sun, Shengzhu Shi, Zhichang Guo, Boying Wu

rag benchmarks safety

arXiv Machine Learning

Jul 7

Binary Iterative Method for Non-targeted Adversarial Attack

arXiv:2607. 04145v1 Announce Type: new Abstract: Adversarial attacks guide and provide additional training and test data for both adversarial training and adversarial robustness validation, and expose the 'piecewise linearity' of deep learning based models.

By Naman Goyal, Milan Chaudhari

safety

One email a morning, machine-written

Related stories

Improving Adversarial Transferability on Vision-Language Pre-training Models via Surrogate-Specific Bias Correction

DeepDefense: Robust Learning via Layer-Wise Gradient-Feature Alignment

Merge Now, Regret Later: The Hidden Cost of Model Merging Is Adversarial Transferability

Investigating Adversarial Robustness of Multi-modal Large Language Models

Latent Geometric Chords for Query-Efficient Decision-Based Adversarial Attacks

Binary Iterative Method for Non-targeted Adversarial Attack