Vision-language-action (VLA) models, world-action models (WAMs), and offline reinforcement learning methods are rapidly expanding the design space of embodied policies, yet turning these algorithms into reliable robot systems remains constrained by fragmented data formats, training stacks, evaluation protocols, inference runtimes, and embodiment-specific interfaces. We present $\mathrm{FluxVLA}$ Engine, an open, configuration-driven platform that turns heterogeneous embodied-policy components into a reproducible data-to-deployment workflow. Rather than introducing another policy model, $\mathrm{FluxVLA}$ standardizes interfaces for datasets, visual-language and world models, action heads, reward- or advantage-weighted learning, distributed training, simulation evaluation, optimized inference, and robot operators. The engine further integrates compositional dual-arm simulation, scalable automatic data generation, and model-decoupled human-in-the-loop rollout, takeover, correction collection, and reward annotation. For responsive physical execution, it combines Real-Time Chunking (RTC) with accelerated inference backends, lightweight remote GPU serving, and configurable trajectory post-processing. Together, these capabilities connect offline learning, simulation validation, online correction, and real-robot execution through shared and auditable contracts. $\mathrm{FluxVLA}$ therefore targets the engineering bottlenecks separating promising embodied-learning algorithms from reproducible evaluation and dependable deployment. Code is available at https://github.com/FluxVLA/FluxVLA
Yin-Hao Li, Wei-Xin Mao, Zi-Han Lan et al.· 1 citation
Stroke disrupts the brain's ability to process and integrate information over time, yet the underlying molecular mechanisms remain unclear. This study investigates post-stroke alterations in intrinsic neural timescales (INTs)-a measure of regional temporal integration-by combining resting-state fMRI, spatial transcriptomics, and PET-based neurochemical mapping. Fifty acute ischemic stroke patients (within 7 days of onset) and fifty matched healthy controls were examined. Voxel-wise and network-level analyses revealed significantly reduced INTs in temporoparietal and insular cortices, with pronounced network-level impairments in the visual and cerebellar systems. Using data from the Allen Human Brain Atlas, we identified gene expression patterns associated with these disruptions. Genes negatively associated with INT reductions were enriched for synaptic, mitochondrial, and neurodevelopmental pathways, while positively associated genes reflected immune signaling and nuclear transport. INT alterations also correlated with excitatory and inhibitory neuronal signatures and were spatially aligned with GABA-A receptor density. Importantly, INT reductions showed significant correlations with clinical assessments: whole-brain INT correlated negatively with NIHSS (r = -0.41) and positively with MoCA (r = 0.38) and FMA (r = 0.35); SMN INT correlated with FMA motor subscore (r = 0.44); and regional INT correlated with domain-specific NIHSS subscores (neglect: r = -0.42; language: r = -0.38). These findings position INT as a clinically meaningful systems-level correlate of stroke-induced dysfunction and highlight molecular pathways and neurotransmitter systems that may constrain temporal integration and recovery potential.
Shao-Gao Gui, Zhan-Xiang Hu, Yuan-Zhi He et al.· Brain Research· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
