arXiv subjects

Megan Lee

Publications and source records attributed to Megan Lee.

4 recordsLinked to original sources

arXiv2026

RIO: Flexible Real-Time Robot I/O for Cross-Embodiment Robot Learning

Despite recent efforts to collect multi-task, multi-embodiment datasets, to design recipes for training Vision-Language-Action models (VLAs), and to showcase these models on different robot platforms, generalist cross-embodiment robot capabilities remains a largely elusive ideal. Progress is limited by fragmented infrastructure: most robot code is highly specific to the exact setup the user decided on, which adds major overhead when attempting to reuse, recycle, or share artifacts between users. We present RIO (Robot I/O), an open source Python framework that provides flexible, lightweight components for robot control, teleoperation, data formatting, sensor configuration, and policy deployment across diverse hardware platforms and morphologies. RIO provides abstractions that enable users to make any choice and to switch between them, with minimal reconfiguration effort. We validate RIO on VLA deployment workflows across three morphologies (single-arm, bimanual, humanoid) and four hardware platforms with varying grippers and cameras. Using teleoperated data collected with RIO, we fine-tune state-of-the-art VLAs including $\pi_{0.5}$ and GR00T on household tasks such as pick-and-place, folding, and bowl scrubbing. By open sourcing all our efforts, we hope the community can accelerate their pace of robot learning on real-world robot hardware. Additional details at: https://robot-i-o.github.io

cs.RO↗

arXiv2026

ASPEN: Spectral-Temporal Fusion for Cross-Subject Brain Decoding

Cross-subject generalization in EEG-based brain-computer interfaces (BCIs) remains challenging due to individual variability in neural signals. We investigate whether spectral representations offer more stable features for cross-subject transfer than temporal waveforms. Through correlation analyses across three EEG paradigms (SSVEP, P300, and Motor Imagery), we find that spectral features exhibit consistently higher cross-subject similarity than temporal signals. Motivated by this observation, we introduce ASPEN, a hybrid architecture that combines spectral and temporal feature streams via multiplicative fusion, requiring cross-modal agreement for features to propagate. Experiments across six benchmark datasets reveal that ASPEN is able to dynamically achieve the optimal spectral-temporal balance depending on the paradigm. ASPEN achieves the best unseen-subject accuracy on three of six datasets and competitive performance on others, demonstrating that multiplicative multimodal fusion enables effective cross-subject generalization.

cs.LG↗

arXiv2021

Dynamics and Rheology of Ring-Linear Blend Semidilute Solutions in Extensional Flow: Single Molecule Experiments

Ring polymers exhibit unique flow properties due to their closed chain topology. Despite recent progress, we have not yet achieved a full understanding of the nonequilibrium flow behavior of rings in nondilute solutions where intermolecular interactions greatly influence chain dynamics. In this work, we directly observe the dynamics of DNA rings in semidilute ring-linear polymer blends using single molecule techniques. We systematically investigate ring polymer relaxation dynamics from high extension and transient and steady-state stretching dynamics in planar extensional flow for a series of ring-linear blends with varying ring fraction. Our results show multiple molecular sub-populations for ring relaxation in ring-linear blends, as well as large conformational fluctuations for rings in steady extensional flow, even long after the initial transient stretching process has subsided. We further quantify the magnitude and characteristic timescales of ring conformational fluctuations as a function of blend composition. Interestingly, we find that the magnitude of ring conformational fluctuations follows a non-monotonic response with increasing ring fraction, first increasing at low ring fraction and then substantially decreasing at large ring fraction in ring-linear blends. A unique set of ring polymer conformations are observed during the transient stretching process, which highlights the prevalence of molecular individualism and supports the notion of complex intermolecular interactions in ring-linear polymer blends. Together with results from molecular simulations, our results suggest that ring conformational fluctuations arise due to ring-linear threading and intermolecular hydrodynamic interactions (HI). Taken together, our results provide a new molecular understanding of ring polymer dynamics in ring-linear blends in nonequilibrium flow.

cond-mat.soft↗

arXiv2019

Unexpected entanglement dynamics in semidilute blends of supercoiled and ring DNA

Blends of polymers of different topologies, such as ring and supercoiled, naturally occur in biology and often exhibit emergent viscoelastic properties coveted in industry. However, due to their complexity, along with the difficulty of producing polymers of different topologies, the dynamics of topological polymer blends remains poorly understood. We address this void by using both passive and active microrheology to characterize the linear and nonlinear rheological properties of blends of relaxed circular and supercoiled DNA. We characterize the dynamics as we vary the concentration from below the overlap concentration c* to above (0.5c* to 2c*). Surprisingly, despite working at the dilute-semidilute crossover, entanglement dynamics, such as shear-thinning, elastic plateaus, and multiple relaxation modes, emerge. Finally, blends exhibit an unexpected sustained elastic response to nonlinear strains not previously observed even in well-entangled linear polymer solutions.

cond-mat.soft↗