Wei Zhao

Tripwire: Triggering Aligned Refusal via Statistically Certified Safety Neurons

Wei Zhao, Zhe Li, Peixin Zhang, et al.

How Do VLMs Behave When Blind or Misled? Behavioral Evaluation of VLMs on Scientific Figures

Paul Osemudiame Oamen, Owusu-Banahene Osei, Ananya Mukherjee, et al.

Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA

Mind Lab, :, Vin Bo, et al.

Looped World Models

Hongyuan Adam Lu, Z. L. Victor Wei, Qun Zhang, et al.

Beyond Accuracy: Community Perspectives on Machine Translation

Yujun Wang, Ehud Reiter, Shimei Pan, et al.

Empowering Heterogeneous Graph Foundation Models via Decoupled Relation Alignment

Ziyu Zheng, Yaming Yang, Zhe Wang, et al.

Are Independently Estimated View Uncertainties Comparable? Unified Routing for Trusted Multi-View Classification

Yilin Zhang, Cai Xu, Haishun Chen, et al.

What Really Controls Temporal Reasoning in Large Language Models: Tokenisation or Representation of Time?

Gagan Bhatia, Ahmad Muhammad Isa, Maxime Peyrard, et al.

This paper introduces MultiTempBench, a benchmark for testing how well large language models handle temporal reasoning (like date arithmetic and timezone conversion) across five languages and different calendar systems. The researchers discovered that how dates are broken into tokens (small text pieces) is crucial for low-resource languages, while high-resource languages rely more on learning temporal patterns, revealing that tokenization quality is a key bottleneck limiting AI's ability to reason about time.