影响力指数
论文质量、代表作、近期表现、广度与样本量置信度综合计算
57.16/100
前 5.6%
全站排名 #3,625
发表论文14 篇
平均评分
年均产出4.7 篇/年
Krishnamurthy Dj Dvijotham
研究方向
privacy · Adversarial attacks · security · deep learning · optimization · control theory · machine learning
15
No, of Course I Can! Deeper Fine-Tuning Attacks That Bypass Token-Level Safety Mechanisms
ICLR 2026Poster
通讯19
Through the Stealth Lens: Attention-Aware Defenses Against Poisoning in RAG
ICLR 2026Rejected
26
Norm-Bounded Low-Rank Adaptation
ICLR 2026Rejected
二作14
VeriGuard: Enhancing LLM Agent Safety via Verified Code Generation
ICLR 2026Withdrawn
23
Indirect Prompt Injections: Are Firewalls All You Need, or Stronger Benchmarks?
ICLR 2026Rejected
13
Malice in Agentland: Down the Rabbit Hole of Backdoors in the AI Supply Chain
ICLR 2026Rejected
23
No, of Course I Can! Deeper Fine-Tuning Attacks That Bypass Token-Level Safety Mechanisms
NeurIPS 2025Rejected
通讯20
DoomArena: A framework for Testing AI Agents Against Evolving Security Threats
COLM 2025Poster
通讯37
BigDocs: An Open Dataset for Training Multimodal Models on Document and Code Tasks
ICLR 2025Poster