Taehyun Kim

In this review, we introduce the latest discoveries in the interaction of lasers with various functional polymers, discussing energy delivery mechanisms and their influence on structure modifications and property changes.

The environmental and therapeutic application of genetically engineered microorganisms necessitates the development of robust, irreversible biocontainment systems. In this study, we present an eEGM (editing-driven essential gene multiplex inactivation) module that utilizes CRISPR-mediated cytidine base editing to induce permanent self-killing via a single transient induction. By targeting the start codons of essential genes, we achieved an irreversible translational blockade that avoids the fitness costs associated with basal toxicity in nuclease-based systems. Multiplexed targeting of non-redundant essential loci (holA, ftsB, and dfp) yielded escape frequencies at or below the NIH guideline criterion (10−8) within 1 h of pulse induction. Furthermore, the eEGM system exhibited robust functional orthogonality and portability across laboratory, industrial, and therapeutic Escherichia coli strains, including MG1655, W3110, and Nissle 1917, without detectable interference with heterologous protein expression. This work establishes base editing as a cleavage-free CRISPR effector for pulse-activated, irreversible biocontainment and provides a practical framework for safer deployment of engineered microbes.

Neural post-processing has been proposed as a lightweight route to enhance variational quantum eigensolvers by learning how to reweight measurement outcomes. In this work, we identify three general desiderata for such data-driven neural post-processing -- (i) self-contained training without prior knowledge, (ii) polynomial resources, and (iii) variational consistency -- and show that current approaches, such as diagonal non-unitary post-processing (DNP), cannot satisfy these requirements simultaneously. The obstruction is intrinsic: with finite sampling, normalization becomes a statistical bottleneck, and support mismatch between numerator and denominator estimators can render the empirical objective ill-conditioned and even sub-variational. Moreover, to reproduce the ground state with constant-depth ansatzes or with linear-depth circuits forming unitary 2-designs, the required reweighting range (and hence the sampling cost) grows exponentially with the number of qubits. Motivated by this no-go result, we develop a normalization-free alternative, the unitary variational quantum-neural hybrid eigensolver (U-VQNHE). U-VQNHE retains the practical appeal of a learnable diagonal post-processing layer while guaranteeing variational safety, and numerical experiments on transverse-field Ising models demonstrate improved accuracy and robustness over both VQE and DNP-based variants.

Trapped ions naturally host multiple motional modes alongside long-lived spin qubits, providing a scalable multimode bosonic register. Efficiently characterizing such bosonic registers requires the ability to access many motional modes with limited spin resources. Here we introduce a single-spin, multimode measurement primitive using dispersive shifts in the far-detuned multimode Jaynes-Cummings interaction. We implement a Ramsey sequence that maps phonon-number-dependent phases onto the spin, thereby realizing a multimode spin-dependent rotation (SDR). We also introduce a selective-decoupling scheme that cancels the phase induced by the carrier AC-Stark shift while preserving the phonon-number-dependent phase induced by the dispersive shift. Using this SDR-based Ramsey sequence on a single trapped ion, we experimentally extract two-mode Fock-state distributions, perform parity-based filtering of two-mode motional states, and realize a nondestructive single-shot measurement of a single-mode Fock state via repeated filtering steps.

Mobile vision assist applications offer critical support for blind and low vision individuals in performing everyday tasks, yet few studies have examined their real-world usability. This study explored two widely used applications - SeeingAI and Sullivan+-through observations and focus group interviews with 17 blind and low vision participants in South Korea performing personal hygiene, shopping, and restroom navigation tasks. Thematic analysis revealed four key themes: (1) Variable Information Preferences Across Contexts, where participants required different information granularity based on situational needs and privacy concerns; (2) Interaction Efficiency and Operational Guidance Challenges, with difficulties from manual capture requirements, lack of camera positioning feedback, and processing delays; (3) Application and System Integration and Compatibility Barriers, highlighting disruptions from inter-application switching and poor accessibility tool integration; and (4) Technology Access and Knowledge Disparities, revealing how educational opportunities, financial constraints, and information dissemination pathways create unequal access within the blind and low vision community. Design recommendations include context-aware information delivery, conversational interfaces, hybrid recognition modes, seamless system integration, and device-agnostic functionality. These findings inform development of more inclusive mobile vision assist applications that support independent living while addressing systemic barriers to equitable technology access.