← Search

Chen Wen

1 accepted papers

2026

Attention Hijacking: Backdooring Text Dataset Distillation via Semantic Anchors

ICML 2026poster

Dataset Distillation (DD) has emerged as a promising technique for compressing large-scale datasets into compact synthetic sets while preserving model performance. However, the security implications of this paradigm, particularly within the Transformer-based text classification domain, remain undere…

Cited by 0SourceScholar