pith. sign in

Geng Hong

Identifiers

  • name variant Geng Hong 0.60 · backfill

Papers (9)

  1. AgentCyberRange: Benchmarking Frontier AI Systems in Realistic Cyber Ranges cs.CR · 2026 · author #12
  2. The Emergence of Autonomous Penetration Capabilities in Large Language Model-Powered AI Systems cs.CR · 2026 · author #8
  3. Unified Safe In-context Image Generation in Multimodal Diffusion Transformers via Restricting Unsafe Information Flows cs.CV · 2026 · author #4
  4. AgentGuard: An Attribute-Based Access Control Framework for Tool-Use LLM-Based Agent cs.CR · 2026 · author #6
  5. Broken Memories: Detecting and Mitigating Memorization in Diffusion Models with Degraded Generations cs.CV · 2026 · author #5
  6. From Anchors to Supervision: Memory-Graph Guided Corpus-Free Unlearning for Large Language Models cs.CL · 2026 · author #4
  7. ReasoningGuard: Safeguarding Large Reasoning Models with Inference-time Safety Aha Moments cs.CL · 2025 · author #4
  8. PRISON: Unmasking the Criminal Potential of Large Language Models cs.CR · 2025 · author #2
  9. Think Twice Before You Act: Enhancing Agent Behavioral Safety with Thought Correction cs.AI · 2025 · author #4

Mentions

  • 2506.16150 #2 · arxiv_oai · confidence 0.70 Geng Hong
  • 2606.14295 #12 · arxiv_oai · confidence 0.70 Geng Hong
  • 2606.13079 #8 · arxiv_oai · confidence 0.70 Geng Hong
  • 2606.06875 #4 · arxiv_oai · confidence 0.70 Geng Hong
  • 2605.28071 #6 · arxiv_oai · confidence 0.70 Geng Hong
  • 2505.11063 #4 · arxiv_oai · confidence 0.70 Geng Hong
  • 2605.22050 #5 · arxiv_oai · confidence 0.70 Geng Hong

Frequent Coauthors