ReShift is a reasoning-level backdoor framework for VLMs that uses poisoned data construction and joint optimization to shift CoT trajectories on trigger while preserving surface coherence.
arXiv preprint arXiv:2505.16916 (2025)
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
years
2026 4verdicts
UNVERDICTED 4representative citing papers
ThinkDeception introduces MLLMs, a multimodal CoT dataset, and VAC-GRPO progressive RL to convert deception detection into interpretable reasoning and claims new SOTA accuracy plus rationale quality.
Sparse autoencoders identify shared latent features across diverse backdoor attacks in LLMs that enable unified detection via classifiers, causal control via steering, and mitigation via ablation fine-tuning.
TIGS detects backdoor-induced attention collapse in LLMs and applies content-aware tail-risk screening plus intrinsic geometric smoothing to suppress attacks while preserving normal performance.
citing papers explorer
-
ReShift: Aha-Moment-Driven Reasoning-Level Backdoor Attacks on Vision-Language Models
ReShift is a reasoning-level backdoor framework for VLMs that uses poisoned data construction and joint optimization to shift CoT trajectories on trigger while preserving surface coherence.
-
ThinkDeception: A Progressive Reinforcement Learning Framework for Interpretable Multimodal Deception Detection
ThinkDeception introduces MLLMs, a multimodal CoT dataset, and VAC-GRPO progressive RL to convert deception detection into interpretable reasoning and claims new SOTA accuracy plus rationale quality.
-
Shared Latent Structures Enable Unified Backdoor Detection and Mitigation in LLMs
Sparse autoencoders identify shared latent features across diverse backdoor attacks in LLMs that enable unified detection via classifiers, causal control via steering, and mitigation via ablation fine-tuning.
-
Defusing the Trigger: Plug-and-Play Defense for Backdoored LLMs via Tail-Risk Intrinsic Geometric Smoothing
TIGS detects backdoor-induced attention collapse in LLMs and applies content-aware tail-risk screening plus intrinsic geometric smoothing to suppress attacks while preserving normal performance.