Shahar Katz

14 papers A* 2Journal 8Unranked 4
YearRankTypeTitle / Venue / Authors
2026 A* conf
AlignTree: Efficient Defense Against LLM Jailbreak Attacks.
AAAI
Gil Goren, Shahar Katz, Lior Wolf
2026 — conf
Safeguarding Language Models via Self-Destruct Trapdoor.
EACL (Volume 1: Long Papers)
Shahar Katz, Bar Alon, Ariel Shaulov, Lior Wolf, Mahmood Sharif
2026 J jnl
TensorLens: End-to-End Transformer Analysis via High-Order Attention Tensors.
CoRR
Ido Atad, Itamar Zimerman, Shahar Katz, Lior Wolf
2025 J jnl
AlignTree: Efficient Defense Against LLM Jailbreak Attacks.
CoRR
Gil Goren, Shahar Katz, Lior Wolf
2025 J jnl
Detecting and Pruning Prominent but Detrimental Neurons in Large Language Models.
CoRR
Ameen Ali, Shahar Katz, Lior Wolf, Ivan Titov
2025 J jnl
Execution Guided Line-by-Line Code Generation.
CoRR
Boaz Lavon, Shahar Katz, Lior Wolf
2025 — conf
Reversed Attention: On The Gradient Descent Of Attention Layers In GPT.
NAACL (Long Papers)
Shahar Katz, Lior Wolf
2025 — conf
Segment-Based Attention Masking for GPTs.
ACL (1)
Shahar Katz, Liran Ringel, Yaniv Romano, Lior Wolf
2024 A* conf
Backward Lens: Projecting Language Model Gradients into the Vocabulary Space.
EMNLP
Shahar Katz, Yonatan Belinkov, Mor Geva, Lior Wolf
2024 J jnl
Backward Lens: Projecting Language Model Gradients into the Vocabulary Space.
CoRR
Shahar Katz, Yonatan Belinkov, Mor Geva, Lior Wolf
2024 J jnl
Reversed Attention: On The Gradient Descent Of Attention Layers In GPT.
CoRR
Shahar Katz, Lior Wolf
2024 J jnl
Segment-Based Attention Masking for GPTs.
CoRR
Shahar Katz, Liran Ringel, Yaniv Romano, Lior Wolf
2023 J jnl
Interpreting Transformer's Attention Dynamic Memory and Visualizing the Semantic Information Flow of GPT.
CoRR
Shahar Katz, Yonatan Belinkov
2023 — conf
VISIT: Visualizing and Interpreting the Semantic Information Flow of Transformers.
EMNLP (Findings)
Shahar Katz, Yonatan Belinkov