arXiv:2606. 19297v1 Announce Type: new Abstract: Embodied Vision-Language-Action (VLA) models are typically obtained by fine-tuning powerful pretrained VLMs on robotics data, yet it is unclear how much commonsense and factual knowledge they retain after adaptation.
By Nikita Kachaev, Andrey Moskalenko, Matvey Skripkin, Nikita Kurlaev, Daria Pugacheva, Albina Burlova, Mikhail Kolosov, Denis Shepelev, Andrey Kuznetsov, Elena Tutubalina, Aleksandr I. Panov, Alexey K. Kovalev, Vlad Shakhuro
Embodied Vision-Language-Action (VLA) models are typically obtained by fine-tuning powerful pretrained VLMs on robotics data, yet it is unclear how much commonsense and factual knowledge they retain after adaptation. Failures on knowledge-sensitive tasks are ambiguous, conflating missing knowledge with poor generalization of low-level control.
arXiv:2603. 06001v2 Announce Type: replace-cross Abstract: Vision-Language-Action (VLA) models enable robots to perform manipulation tasks directly from natural language instructions and are increasingly viewed as a foundation for generalist robotic policies.
By Ninghao Zhang, Bin Zhu, Shijie Zhou, Jingjing Chen
Latent‑Lens is a lightweight codec that enables a frozen vision‑language model (VLM) to communicate visual information to a frozen small language model (SLM) directly in latent space, avoiding the lossy bottleneck of discrete text. The codec uses a domain encoder to project VLM hidden states into the SLM’s embedding space and an attention decoder (a LoRA) that teaches the SLM to attend to these visual prefixes, training only about 1.6 M parameters (0.2 % of the combined model size). On visual question answering tasks (CLEVR, GQA) and image captioning (Flickr8k), Latent‑Lens significantly outperforms text‑based communication with the same LoRA budget, achieving up to +69.8 pp improvement on VQA and +18.8 pp on captioning.
By Mo Shahdloo, Faegheh Sardari, Mathew Salvaris
arXiv:2606. 16898v1 Announce Type: cross Abstract: Detecting unanswerable user queries remains essential for the reliable deployment of real-world embodied agents.
By Dongbin Na, Chanwoo Kim, Giyun Choi, Dooyoung Hong
The paper introduces V-JEPA-LM, a video‑language model that leverages the dynamic representations learned by V‑JEPA video encoders. By applying counterfactual supervision—pixel transformations that create answer‑balanced video variants—the model learns to read motion dynamics such as speed, direction, and order, rather than merely retrieving static attributes. Trained on 4.53 M public samples with 32 visual tokens per frame, V‑JEPA‑LM outperforms existing open‑source video LLMs across eleven benchmarks, with a 4 B model achieving a 72.1 average score.
By Keuntae Kim, Dongjun Kim, Yong Suk Choi