arXiv Machine Learning August 21, 2026 By Kaizhen Tan, Xin Xu, Siru Tao, Yixiao Li, Hanzhe Hong, Yang Feng, Heqing Du

Beyond Multimodal Alignment: Certifying Physical Language through Response Substitution and Ordered Execution

Read the original on arXiv Machine Learning →

arXiv:2608. 19492v1 Announce Type: new Abstract: World models increasingly treat compact multimodal representations as interfaces between perception and physical interaction, yet existing probes do not establish whether different sensors carry the same executable meaning or whether that meaning survives a new action composition.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv Machine Learning.

arXiv Machine Learning

Aug 11

When Do Task Vectors Interfere? Mapping the Validity Boundaries of Weight-Space Composition

arXiv:2608. 09490v1 Announce Type: new Abstract: Task arithmetic treats fine-tuning displacements as composable directions in weight space, yet it remains unclear when parameter addition reflects predictable changes in model function.

By Chencheng Zhu, Xiaoyang Li, Taotao Cai

llms fine-tuning safety

arXiv AI

Aug 19

Where a New Concept Must Enter: Entry Point Gates Cross-Task Usability in Unified Multimodal Models

The paper investigates how new concepts can be integrated into unified multimodal models (UMMs) by separating generation and understanding objectives through a novel visual entity bound to a single task direction. Experiments show that the effectiveness of cross‑task usability depends on where the concept is injected into the shared computation, with a mid‑stack alignment objective achieving high concept acquisition with minimal loss to overall performance. The study highlights that unified weights alone are insufficient; the two directions must share a semantic format at the entry point for efficient concept integration.

By Zongyang Qiu, Yihan Wu, Kaixuan Fan, Bo Li, Hui Xiong

diffusion multimodal safety

arXiv Machine Learning

Jun 15

PhysVLA: Towards Physically-Grounded VLA for Embodied Robotic Manipulation

arXiv:2606. 13886v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models excel at mapping visual inputs and natural language instructions directly to robotic control policies.

By Namai Chandra, Shriram Damodaran, Lin Wang

robotics fine-tuning multimodal

arXiv Computer Vision

Sep 23

Sample, Simulate, Select: Physics-in-the-Loop Text-to-Motion for Humanoids Without Training

arXiv:2609.26420v1 Announce Type: cross Abstract: Text-to-motion models generate plausible human motion but do not model a robot's dynamics; whole-body tracking controllers execute robot references r...

By Raphael Memmesheimer, Sven Behnke

robotics

arXiv AI

Jul 22

Operational Proto-Introspection in Looped Language Models: Process-Quality Taps, Executable Branching, and the Readout-Control Boundary

arXiv:2607. 18553v1 Announce Type: cross Abstract: Can a language model read the quality of ongoing computation, and can an external intervention turn that readout into better outcomes?

By Jan Kirin

llms fine-tuning

arXiv Computer Vision

Sep 15

What Makes an Efficient VLA? Navigating Action-Head Design, Scaling, and Latency

arXiv:2609.13984v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models combine a pretrained vision encoder, a language backbone, and an action head, but their relative contribution has...

By Luoyang Sun, Guoyang Xia, Fengfa Li, Lei Ren, Xinyu Cui, Haifeng Zhang, Fangxiang Feng, Kaike Zhang, Kun Zhan, Yan Xie, Jun Wang, Cheng Deng

llms diffusion multimodal safety