Search for: All records

Zeroth-order fine-tuning eliminates explicit back-propagation and reduces memory overhead for large language models (LLMs), making it a promising approach for on-device fine-tuning tasks. However, existing memory-centric accelerators fail to fully leverage these benefits due to inefficiencies in balancing bit density, compute-in-memory capability, and endurance-retention trade-off. We present a reliability-aware, analog multi-level-cell (MLC) eDRAM-RRAM compute-in-memory (CIM) solution co-designed with zeroth-order optimization for language model fine-tuning. An RRAM-assisted eDRAM MLC programming scheme is developed, along with a process-voltage-temperature (PVT)-robust, large-sensing-window time-to-digital converter (TDC). The MLC-eDRAM integrating two-finger MOM provides 12× improvement in bit density over state-of-the-art MLC design. Another 5× density and 2× retention benefits are gained by adopting BEOL In2O3 FETs. 

We developed a physics-informed machine learning platform that predicts stress–strain curves of 3D-printed thermoplastics from ink formulations, enabling virtual experimentation and rapid identification of optimal materials in vast chemical spaces. 

In nature, structural and functional materials often form programmed three-dimensional (3D) assembly to perform daily functions, inspiring researchers to engineer multifunctional 3D structures. Despite much progress, a general method to fabricate and assemble a broad range of materials into functional 3D objects remains limited. Herein, to bridge the gap, we demonstrate a freeform multimaterial assembly process (FMAP) by integrating 3D printing (fused filament fabrication (FFF), direct ink writing (DIW)) with freeform laser induction (FLI). 3D printing performs the 3D structural material assembly, while FLI fabricates the functional materials in predesigned 3D space by synergistic, programmed control. This paper showcases the versatility of FMAP in spatially fabricating various types of functional materials (metals, semiconductors) within 3D structures for applications in crossbar circuits for LED display, a strain sensor for multifunctional springs and haptic manipulators, a UV sensor, a 3D electromagnet as a magnetic encoder, capacitive sensors for human machine interface, and an integrated microfluidic reactor with a built-in Joule heater for nanomaterial synthesis. This success underscores the potential of FMAP to redefine 3D printing and FLI for programmed multimaterial assembly. 

Inbred mice used for biomedical research display an underdeveloped immune system compared with adult humans, which is attributed in part to the artificial laboratory environment. Despite representing a central component of adaptive immunity, the impact of the laboratory environment on the B cell compartment has not been investigated in detail. Here, we performed an in-depth examination of B cells following rewilding, the controlled release of inbred laboratory mice into an outdoor enclosure. In rewilded mice, we observed B cells in circulation with increased signs of maturation, alongside heightened germinal center responses within secondary lymphoid organs. Rewilding also expanded B cells in the gut, which was accompanied by elevated systemic levels of immunoglobulin G (IgG) and IgM antibodies reactive to the microbiota. Our findings indicate that exposing laboratory mice to a more natural environment enhances B cell development to better reflect the immune system of free-living mammals. 

Rapid analysis of materials characterization spectra is pivotal for preventing the accumulation of unwieldy datasets, thus accelerating subsequent decision-making. However, current methods heavily rely on experience and domain knowledge, which not only proves tedious but also makes it hard to keep up with the pace of data acquisition. In this context, we introduce a transferable Vision Transformer (ViT) model for the identification of materials from their spectra, including XRD and FTIR. First, an optimal ViT model was trained to predict metal organic frameworks (MOFs) from their XRD spectra. It attains prediction accuracies of 70%, 93%, and 94.9% for Top-1, Top-3, and Top-5, respectively, and a shorter training time of 269 seconds (∼30% faster) in comparison to a convolutional neural network model. The dimension reduction and attention weight map underline its adeptness at capturing relevant features in the XRD spectra for determining the prediction outcome. Moreover, the model can be transferred to a new one for prediction of organic molecules from their FTIR spectra, attaining remarkable Top-1, Top-3, and Top-5 prediction accuracies of 84%, 94.1%, and 96.7%, respectively. The introduced ViT-based model would set a new avenue for handling diverse types of spectroscopic data, thus expediting the materials characterization processes.