aiqbench analysis
This paper introduces $N_0$-VTLA, a foundation model that integrates large-scale tactile data into vision-based backbones for contact-rich robotic manipulation. By employing staged pre-training and an offline reinforcement learning method called ALTER, the system achieves significant performance gains on real-world tasks involving deformable objects without requiring continuous online interaction. The work represents a substantive step forward in scaling multimodal robot policies with latent tactile tokens to handle fine-grained physical interactions.
$N_0$-VTLA: Scaling Vision-Tactile-Language-Action Model with Latent Tactile Tokens
arXiv:2607.23782v1 Announce Type: new Abstract: We present $N_0$-VTLA, a vision-tactile-language-action (VTLA) foundation model capable of (1) fine-grained contact-rich manipulation with tactile perception and tactile-feedback control, and (2) offline policy improvement from stored deployment data. Building on current vision-based backbones, we propose a training recipe for tactile integration consisting of visuo-tactile pre-training, staged tactile-pathway integration, and advantage-conditioned offline policy improvement. During pre-training, the policy learns broad contact priors from NeoData, our large-scale visuo-tactile robot dataset; to our knowledge, $N_0$-VTLA is the first VTLA model pretrained on tactile data at scale. During post-training, we augment the policy with a predictive tactile pathway that distills the contact patterns learned at scale into the fine motion adjustments required by downstream tactile-centric manipulation. For offline policy improvement, we introduce ALTER, an advantage-conditioned offline reinforcement learning method that converts relative progress and trajectory-event comparisons into binary advantage labels for policy training on a fixed deployment corpus, further improving task-specific learning on contact-rich skills such as deformable object manipulation. Across contact-rich benchmarks, $N_0$-VTLA outperforms strong baselines by wide margins: it wins all nine real-robot NeoReal tasks and reaches 63.8% mean success on a twenty-task simulation suite, against 44.0% for the strongest baseline. $N_0$-VTLA policies trained with ALTER reach 75-95% success on three long-horizon real-robot tasks. These results lay a foundation for versatile tactile-driven manipulation policies.
Abstract:We present $N_0$-VTLA, a vision-tactile-language-action (VTLA) foundation model capable of (1) fine-grained contact-rich manipulation with tactile perception and tactile-feedback control, and (2) offline policy improvement from stored deployment data. Building on current vision-based backbones, we propose a training recipe for tactile integration consisting of visuo-tactile pre-training, staged tactile-pathway integration, and advantage-conditioned offline policy improvement. During pre-training, the policy learns broad contact priors from NeoData, our large-scale visuo-tactile robot dataset; to our knowledge, $N_0$-VTLA is the first VTLA model pretrained on tactile data at scale. During post-training, we augment the policy with a predictive tactile pathway that distills the contact patterns learned at scale into the fine motion adjustments required by downstream tactile-centric manipulation. For offline policy improvement, we introduce ALTER, an advantage-conditioned offline reinforcement learning method that converts relative progress and trajectory-event comparisons into binary advantage labels for policy training on a fixed deployment corpus, further improving task-specific learning on contact-rich skills such as deformable object manipulation. Across contact-rich benchmarks, $N_0$-VTLA outperforms strong baselines by wide margins: it wins all nine real-robot NeoReal tasks and reaches 63.8% mean success on a twenty-task simulation suite, against 44.0% for the strongest baseline. $N_0$-VTLA policies trained with ALTER reach 75-95% success on three long-horizon real-robot tasks. These results lay a foundation for versatile tactile-driven manipulation policies.
Robotics (cs.RO)
arXiv:2607.23782 [cs.RO]
(or arXiv:2607.23782v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2607.23782 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 306 words
Related coverage
TacClaw-Infinity: A Large-Scale Vision-Tactile-Language-Action Dataset for Dexterous Manipulation
arXiv · Oct 15, 2026
Spark-0: Scaling Spatially Aligned Tactile Pretraining for Robot Dexterous Manipulation
arXiv · Oct 16, 2026
STAR: Sparse Tactile Representation Learning in Vision-Tactile-Language-Action Models for Dexterous Manipulation
arXiv · Sep 14, 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · Oct 13, 2026