4 papers
Teaching Vision-Language Models to Use the Scale They Are Given: Label-Free Equivariance Training for Metric Physical Reasoning
Kaizhen Tan, Yang Feng, Heqing Du +3
Metric questions about video require vision-language models to use supplied real-world references to convert visual measurements into physical units. Yet we find that current model…
Beyond Multimodal Alignment: Certifying Physical Language through Response Substitution and Ordered Execution
Kaizhen Tan, Xin Xu, Siru Tao +4
World models increasingly treat compact multimodal representations as interfaces between perception and physical interaction, yet existing probes do not establish whether different…
What Can Latent World Models Know? Physical Parameter Identifiability in Multimodal Predictive Representations
Kaizhen Tan, Xin Xu, Siru Tao +4
A central premise of latent world models is that predicting the future forces a representation to internalize the physics of its environment. Which physical quantities does a train…
How Many Bits Can an Adapter Write? Measuring the Capacity and Memorization of Parameter-Efficient Fine-Tuning
Kaizhen Tan, Heqing Du, Yang Feng
A LoRA adapter is a few megabytes that almost everyone treats as a skill rather than a record of the data behind it. We put that assumption on a scale. Extending compression-based…