Tal Haklay

4 papers · 2024–2025 · 3 conferences · across top CS/AI conferences

Achievements

🧭 Keyword Pioneer 🌍 Conference Polyglot (3) 🐝 Cross-Pollinator (4) 🌉 Interdisciplinary Bridge 👥 Mega-Team (23)

Conferences

ICLR (2) ACL (1) ICML (1)

Top co-authors

Yonatan Belinkov (4) David Bau (4) Hadas Orgad (2) Nikhil Prakash (2) Aaron Mueller (2) Adam Belfki (1) Arnab Sen Sharma (1) Sarah Wiegreffe (1) Kevin Meng (1) Evan Hernandez (1)

Keywords

language model (1) mechanistic interpretability (1) language model interpretability (1) circuit discovery (1) edge attribution patching (1) position-aware circuit (1) model mechanism analysis (1)

Papers

Position-aware Automatic Circuit Discovery ACL 2025

MIB: A Mechanistic Interpretability Benchmark ICML 2025

Linearity of Relation Decoding in Transformer Language Models ICLR 2024

Fine-Tuning Enhances Existing Mechanisms: A Case Study on Entity Tracking ICLR 2024