COLM 2026 publications
12 Jul 2026

Papers at COLM 2026 (San Francisco, CA).

• Hallucination Self-Play: Bootstrapping Reinforced Detector via Evolved Generator

Linguistics colloquium with Marzena
09 Jul 2026

Marzena is giving a talk about her work on characterizing AI in writing as part of the linguistics colloquium: https://www.sfu.ca/linguistics/events/2026/Karpinska.html

ACL 2026 publications
02 Jul 2026

Papers at ACL 2026 (San Diego, CA).

• Sun July 5 14:00-15:30 Find Your Optimal Teacher: Personalized Data Synthesis via Router-Guided Multi-Teacher Distillation - (Oral Session B: Language Models 2, Harbor H-I)
• Sun July 5 16:00-17:30 AI use in American newspapers is widespread, uneven, and rarely disclosed - (Oral Session C: Computational social science, cultural analytics, and NLP for social good 2, Harbor G)
• Sun July 5 16:00-17:30 Quantifying the robustness of retrieval-augmented language models against spurious features in grounding data - (Poster Session C-90, Coronado)
• Mon July 6 18:00-19:30 PR-XAI: PageRank-Based Feature Attribution for Transformers - (Virtual Presentations 4)

New NLP faculty at SFU
06 Apr 2026

We are happy to have welcome Marzena Karpinska and Linyi Li to the SFU natural language group.

ACL 2025 publications
27 Jul 2025

Papers at ACL 2025.

• ViGiL3D: A Linguistically Diverse Dataset for 3D Visual Grounding
• Generalized Attention Flow: Feature Attribution for Transformer Models via Maximum Flow