Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Kai Wang

Kai Wang

9 indexed papers

Recent (6 mo)
6
With code
0
Influential cites
0
Benchmarked
0

Publications per year

9
26

Top categories

Vision×6NLP×5Crypto×3ML×3AI×2Software Eng.×1Multiagent×1

Frequent co-authors

Hezhen Hu2×
Wangbo Zhao2×
Lanqing Guo2×
Hanwen Jiang2×
Jonathan C. Liu2×
Zhiwen Fan2×

Research Timeline

2026
The Salami Slicing Threat: Exploiting Cumulative Risks in LLM Systems

The paper introduces Salami Slicing Risk, a novel multi-turn jailbreak technique that accumulates harmful intent through numerous low-risk inputs, achieving state-of-the-art attack success rates against major LLMs.

SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaces

The paper introduces SkillSafetyBench, a comprehensive benchmark demonstrating that agent safety failures often stem from adversarial influences within reusable skills and execution environments, rather than just malicious user prompts.

Awakening the Hydra: Stabilizing Multi-Concept Backdoor Injection in Text-to-Image Diffusion Models

The paper proposes Hydra, a framework to stabilize and control the injection of multiple, conflicting backdoor triggers into text-to-image diffusion models, ensuring high attack reliability while maintaining clean generation quality.

Attend to Evidence: Evidence-Anchored Spatial Attention Supervision for Multimodal RLVR

The paper introduces EASE, a method that enhances multimodal Reinforcement Learning with Verifiable Rewards (RLVR) by providing spatial attention supervision anchored to visual evidence, significantly improving visual grounding and reasoning capabilities in VLMs.

Attend to Evidence: Evidence-Anchored Spatial Attention Supervision for Multimodal RLVR

The paper introduces EASE, a method that enhances multimodal Reinforcement Learning with Verifiable Rewards (RLVR) by providing spatial attention supervision anchored to visual evidence, significantly improving visual grounding and reasoning capabilities in VLMs.

HumanNOVA: Photorealistic, Universal and Rapid 3D Human Avatar Modeling from a Single Image

HumanNOVA introduces a photorealistic, universal, and rapid model capable of generating high-quality 3D human avatars from a single input RGB image.

HumanNOVA: Photorealistic, Universal and Rapid 3D Human Avatar Modeling from a Single Image

HumanNOVA introduces a photorealistic, universal, and rapid model capable of generating high-quality 3D human avatars from a single input RGB image.

Token-Operations-Oriented Inference Optimization Techniques for Large Models

This paper proposes a four-layer technical architecture for large model inference optimization, including Multi-model Fusion, Model Optimization, Compute-Model Fusion, and Compute-Network-Model Fusion.

Synthetic Image Oversampling for Rare-Class Representation Learning in Deep Facial Phenotyping

This paper investigates the use of synthetic facial images to supplement training data for recognizing genetic syndromes.

Highlighted terms show continued research focus across papers

Papers

cs.CVEmpiricalRecentOct 16, 2026

Synthetic Image Oversampling for Rare-Class Representation Learning in Deep Facial Phenotyping

Miles Liao, Zhanliang Wang, Hongzhuo Chen, Kai Wang

This paper investigates the use of synthetic facial images to supplement training data for recognizing genetic syndromes.

View →
cs.SEcs.CLSurveyRecent
Jun 18, 2026

Token-Operations-Oriented Inference Optimization Techniques for Large Models

Shiguo Lian, Kai Wang, Zhaoxiang Liu, Wen Liu +21 more

This paper proposes a four-layer technical architecture for large model inference optimization, including Multi-model Fusion, Model Optimization, Compute-Model Fusion, and Compute-Network-Model Fusion…

View →
cs.CVRecentJun 1, 2026

HumanNOVA: Photorealistic, Universal and Rapid 3D Human Avatar Modeling from a Single Image

Hezhen Hu, Wangbo Zhao, Lanqing Guo, Hanwen Jiang +5 more

HumanNOVA introduces a photorealistic, universal, and rapid model capable of generating high-quality 3D human avatars from a single input RGB image.

View →
cs.CVRecentJun 1, 2026

HumanNOVA: Photorealistic, Universal and Rapid 3D Human Avatar Modeling from a Single Image

Hezhen Hu, Wangbo Zhao, Lanqing Guo, Hanwen Jiang +5 more

HumanNOVA introduces a photorealistic, universal, and rapid model capable of generating high-quality 3D human avatars from a single input RGB image.

View →
cs.CVcs.CLRecentMay 29, 2026

Attend to Evidence: Evidence-Anchored Spatial Attention Supervision for Multimodal RLVR

Ruina Hu, Chen Wang, Lai Wei, Jionghao Bai +4 more

The paper introduces EASE, a method that enhances multimodal Reinforcement Learning with Verifiable Rewards (RLVR) by providing spatial attention supervision anchored to visual evidence, significantly…

View →
cs.CVcs.CLRecentMay 29, 2026

Attend to Evidence: Evidence-Anchored Spatial Attention Supervision for Multimodal RLVR

Ruina Hu, Chen Wang, Lai Wei, Jionghao Bai +4 more

The paper introduces EASE, a method that enhances multimodal Reinforcement Learning with Verifiable Rewards (RLVR) by providing spatial attention supervision anchored to visual evidence, significantly…

View →
cs.CRcs.LGMay 19, 2026

Awakening the Hydra: Stabilizing Multi-Concept Backdoor Injection in Text-to-Image Diffusion Models

Kai Wang, Jiale Zhang, Chengcheng Zhu, Chuang Ma +1 more

The paper proposes Hydra, a framework to stabilize and control the injection of multiple, conflicting backdoor triggers into text-to-image diffusion models, ensuring high attack reliability while main…

View →
cs.CRcs.AIcs.CLMay 12, 2026

SkillSafetyBench: Evaluating Agent Safety under Skill-Facing Attack Surfaces

Chang Jin, An Wang, Zeming Wei, Kai Wang +6 more

The paper introduces SkillSafetyBench, a comprehensive benchmark demonstrating that agent safety failures often stem from adversarial influences within reusable skills and execution environments, rath…

View →
cs.CRcs.AIcs.CLApr 13, 2026

The Salami Slicing Threat: Exploiting Cumulative Risks in LLM Systems

Yihao Zhang, Kai Wang, Jiangrong Wu, Haolin Wu +6 more

The paper introduces Salami Slicing Risk, a novel multi-turn jailbreak technique that accumulates harmful intent through numerous low-risk inputs, achieving state-of-the-art attack success rates again…

View →