The sensitive detection of microRNA-21 (miR-21), a key biomarker for various cancers, is crucial for early diagnosis, yet conventional methods often face limitations in sensitivity and operational complexity. Here, we report a label-free biosensor based on a suspended graphene field-effect transistor (GFET) for the direct electrical detection of miR-21. The suspended architecture isolates the graphene channel from substrate-induced interference, resulting in enhanced carrier mobility and reduced electrical noise. After surface functionalization with a specific probe, the GFET demonstrated a clear concentration-dependent response to target miR-21. The binding events were transduced into a monotonic increase in relative resistance (ΔR/R₀) and a positive shift of the Dirac point (V_Dirac), achieving a detection limit in the femtomolar (fM) range. These results establish the suspended GFET as a highly sensitive and robust platform for quantifying nucleic acid biomarkers, holding significant potential for biomedical research and point-of-care diagnostics.

The design of broadband, high-efficiency solar absorbers remains challenging due to the complex and ill-posed inverse mapping from the target optical responses to the physical structures in inverse design optimization. To address this, we propose a joint forward–inverse deep learning framework that enables the rapid and accurate optimization of multilayer metamaterial absorbers. This method integrates an inverse network based on a Modified Swin Transformer with a Multilayer Perceptron forward proxy and performs end-to-end training in a consistency-driven cycle. This strategy reduces the one-to-many ambiguity in inverse design and improves the prediction accuracy, with normalized test mean squared errors of 7.2 × 10⁻⁵ (inverse) and 6.8 × 10⁻⁵ (forward). Using this framework, we optimized an absorber comprising W/SiO₂ hyperbolic metamaterial stacks and TiO₂/SiO₂ anti-reflection coatings, achieving 97.4% average absorptivity across the 400–1750 nm solar spectrum, along with polarization insensitivity and robust wide-angle performance up to 60° incidence. The outdoor solar heating tests showed that the fabricated absorber reaches a peak temperature of 86.3 °C under natural sunlight, with an irradiance peak of about 850 W/m² at noon. This work shows that combining forward and reverse deep learning provides a powerful and scalable paradigm for accelerating the intelligent design of high-performance solar thermal metamaterials.