Jennifer (Meng) Lu

prof_pic.jpg

I’m a Sc.M. student in Computer Science at Brown University, advised by Ellie Pavlick and Carsten Eickhoff. I graduated from Wellesley College with a double major in CS and history.

I am very interested in model interpretability and representation learning. My current research investigates the geometry of model representations, with a focus on understanding how geometric structure supports generalization, whether it can be used to predict model failures, and how these insights can inform more effective training methods.

Outside of research, I enjoy creative writing, chamber music, tennis and reading.

news

Sep 22, 2025 Gave an invited talk at Google Multilingual Group on Paths not Taken with Ruochen!
Aug 23, 2025 Our ongoing work “Mechanisms of In-Context Syntatic Generalization in Language Models” is presented in NEMI 2025 and is accepted to BlackboxNLP 2025!
Aug 20, 2025 My two 1st-author papers both have been accepted to EMNLP 2025 Main Conference! So grateful to my most amazing co-authors and mentors 💐

selected publications

  1. ICML Mech Interp Workshop 2026
    Adversarial Concept Search: Predicting Compositional Errors From Feature Geometry
    Jennifer Meng Lu, Ruochen Zhang, Isabelle Lee, David Alvarez-Melis, Ellie Pavlick, and Naomi Saphra
    2026
  2. EMNLP 2025
    Paths Not Taken: Understanding and Mending the Multilingual Factual Recall Pipeline
    Meng Lu*, Ruochen Zhang*, Carsten Eickhoff, and Ellie Pavlick
    2025
  3. EMNLP 2025
    Pathway to Relevance: How Cross-Encoders Implement a Semantic Variant of BM25
    Meng Lu*, Catherine Chen*, and Carsten Eickhoff
    2025