arXiv AI June 16, 2026 By Zhenyu Yu

InstantForget: Update-Free Backdoor Unlearning with Inference-Time Feature Reset

arXiv:2606. 15730v1 Announce Type: cross Abstract: Backdoor unlearning aims to remove a malicious trigger behavior from a deployed model while preserving clean utility.

Summary generated by The Flow from the publisher's feed. The full article lives at arXiv AI.

multimodal

arXiv AI

Jun 19

Calibration Without Comprehension: Diagnosing the Limits of Fine-Tuning LLMs for Vulnerability Detection in Systems Software

arXiv:2606. 20502v1 Announce Type: cross Abstract: Whether LLMs scoring well on vulnerability benchmarks genuinely reason about security or merely pattern-match on contaminated data remains unresolved.

By Arastoo Zibaeirad, Marco Vieira

llms fine-tuning benchmarks

arXiv Machine Learning

Jun 2

CANARY: Zero-Label Detection of Fine-Tuning Contamination in Language Models

arXiv:2606. 01695v1 Announce Type: new Abstract: Adversaries can implant latent harmful behavior by poisoning as few as 1% of fine-tuning examples.

By Swapnil Parekh

llms fine-tuning safety

arXiv Machine Learning

Jul 21

When Benchmarks Lie: Evaluating Malicious Prompt Classifiers Under True Distribution Shift

arXiv:2602. 14161v2 Announce Type: replace Abstract: Detecting prompt injection, jailbreak attacks, and harmful requests is critical for deploying LLM-based agents safely, yet current evaluation practices in this literature overestimate generalization.

By Max Fomin

llms agents benchmarks safety

arXiv Machine Learning

Aug 4

Z-PEFT: Zero-shot Backdoor Detection in Parameter-Efficient Fine-Tuning via Canonical Spectral Signatures

arXiv:2608. 02271v1 Announce Type: new Abstract: Parameter-Efficient Fine-tuned (PEFT) models are frequently downloaded from open repositories by practitioners.

By Nicola Pitzalis, Donald Shenaj, Giacomo Cignoni, Andrea Cossu, Davide Bacciu, Antonio Carta

fine-tuning safety

arXiv AI

Jul 2

FLAT: Revealing Hidden Latent-Conditioned Backdoor Failures in Federated Learning

arXiv:2508. 04064v2 Announce Type: replace-cross Abstract: Horizontal federated learning (HFL) backdoor audits often summarize model behavior through clean accuracy (CA), mean attack success rate (ASR), or a single known-trigger test.

By Tuan Nguyen, Sze Jue Yang, Khoa D. Doan, Chee Seng Chan, Kok-Seng Wong

multimodal

arXiv AI

Jul 23

Defense Against LLM Backdoors using Critical Neuron Isolation Pruning

arXiv:2607. 19894v1 Announce Type: cross Abstract: Large language models (LLMs) are vulnerable to backdoor attacks, where hidden triggers induce malicious outputs.

By Yuxi Li, Zhibo Zhang, Kailong Wang, Xingshuo Han, Ling Shi, Haoyu Wang

llms fine-tuning efficiency multimodal benchmarks

Related stories

Calibration Without Comprehension: Diagnosing the Limits of Fine-Tuning LLMs for Vulnerability Detection in Systems Software

CANARY: Zero-Label Detection of Fine-Tuning Contamination in Language Models

When Benchmarks Lie: Evaluating Malicious Prompt Classifiers Under True Distribution Shift

Z-PEFT: Zero-shot Backdoor Detection in Parameter-Efficient Fine-Tuning via Canonical Spectral Signatures

FLAT: Revealing Hidden Latent-Conditioned Backdoor Failures in Federated Learning

Defense Against LLM Backdoors using Critical Neuron Isolation Pruning