PPaperPicks

Shahar Katz

7 papers at tracked venues · 4 at CORE A* · active 20242026

Venues

Frequent coauthors

Papers

  1. AlignTree: Efficient Defense Against LLM Jailbreak Attacks
  2. Safeguarding Language Models via Self-Destruct Trapdoor
    EACL 2026 · Shahar Katz
  3. TensorLens: End-to-End Transformer Analysis via High-Order Attention Tensors
  4. Execution Guided Line-by-Line Code Generation
  5. Reversed Attention: On The Gradient Descent Of Attention Layers In GPT
    NAACL 2025 · Shahar Katz
  6. Segment-Based Attention Masking for GPTs
    ACL 2025 · Shahar Katz
  7. Backward Lens: Projecting Language Model Gradients into the Vocabulary Space
    EMNLP 2024 · Shahar Katz