A research team including Ph.D. candidate Young-Seok Lee (first author, co-advised by Professors Jungsuek Oh and Sangwook Nam), from the Wave Fusion Lab, led by Professor Jungsuek Oh, was the sole winner (1st place) of the Best Student Paper Award at the 2026 IEEE Wireless Power Technology Conference & Expo (WPTCE 2026). The event, hosted by IEEE, was held in Halifax, Canada, from July 6th to 9th. Lee delivered an oral presentation titled "DC-to-DC Efficiency Maximization in Wireless Power Transfer with PA and Rectifier Nonlinearities via Semidefinite Relaxation." Following a review by the Technical Program Committee (TPC), he was selected as the singular winner out of 145 papers presented from 26 countries. ​The IEEE WPTCE is the world's largest international conference in the field of wireless power transfer, jointly organized by the IEEE Microwave Theory and Technology Society (MTT-S) and the IEEE Power Electronics Society (PELS). The MTT-S's WPTC (inaugurated in 2013) and the PELS's WoW (inaugurated in 2015) were co-hosted as 'Wireless Power Week' starting in 2018, before officially merging into the single IEEE WPTCE conference in 2023. This marks the first time a researcher from a Korean institution has won first place since the conference's consolidation. This paper departs from conventional methods that evaluate performance based solely on the RF power from the transmitting antenna to the receiving antenna. Instead, it proposes a framework for optimizing end-to-end efficiency—from the DC power consumed by the transmitting power amplifier (PA) to the DC power output from the receiving rectifier. As the first framework of its kind applied to a MIMO transceiver scenario, it is anticipated to provide an upper-bound guideline for end-to-end efficiency in future wireless power transfer systems. Commercially, it is expected to contribute to enhancing the efficiency of multi-antenna-based long-distance wireless charging, 6G ISAC MIMO, as well as drone and space solar wireless power transfer systems. ​Meanwhile, prior to the conference, Lee was selected for the IEEE MTT-S 'Student Travel Grant' (1,000 CAD). Furthermore, an extended study of this paper, titled "An End-to-End DC-to-DC Efficiency Optimization and Design Analysis for MIMO Wireless Power Transfer Systems," was accepted for publication on June 18th in the IEEE Transactions on Microwave Theory and Techniques (TMTT), the most prestigious academic journal in the microwave field. Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57917 Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr...

▲ The Institute of Semiconductor Engineers Summer Conference (2026.07.16) Sunwoo Hong, an M.S.-Ph.D. integrated candidate in the Department of Electrical and Computer Engineering (Advisor: Professor Jerald Yoo), won the Outstanding Paper Award in the poster category at the 2026 Institute of Semiconductor Engineers Summer Conference held at Ananti at Busan Cove on July 16th. The Institute of Semiconductor Engineering Summer Conference is an academic event where researchers from industry, research institutes and academia participate across all semiconductor fields, ranging from semiconductor devices and processing to circuit design and AI semiconductors. This year, it was held in Busan for four days, from July 14th to 17th. Various programs took place, including oral and poster presentations, keynote lectures, special sessions and a Semiconductor Live Demo Competition. The award-winning paper, “Design of a Chopper-Stabilized Capacitor-Coupled Instrumentation Amplifier for Wearable EEG Measurement,” is a study dealing with instrumentation amplifier circuits used in wearable electroencephalogram (EEG) measurement devices. To amplify minute EEG signals at the levels of 10 μV, a chopper-stabilized capacitor-coupled instrumentation amplifier (CS-CCIA) was designed, and its operation was verified through a 180 nm CMOS process simulation. Hong presented the research at the poster session on July 16th, and after undergoing evaluation, won the Outstanding Paper Award in the poster category. In the future, he plans on integrating the designed amplifier with an ADC and a digital backend, developing it into a sensor interface for a wearable EEG measurement system. Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57902 Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr...

AI-Based Reverse Design of Processing Conditions Extends Quantum Dot QLED Lifetime by 40 Times - Improves next-generation display performance by reverse-designing optimal solvent properties for quantum dot processing A technology has been developed that enables artificial intelligence to reverse-design the processing conditions for quantum dot light-emitting diode (QLED) devices after extensive trial and error. When applied to actual devices, the technology doubled efficiency and improved lifetime by more than 40 times, raising expectations that it could accelerate the development of next-generation displays. ▲ (From left) Professor Jeonghun Kwak of the ECE Department at SNU, Professor Jaehoon Lim of Sungkyunkwan University, and Ph.D. candidate Beomsoo Chun of the ECE Department at SNU Seoul National University’s College of Engineering announced that a joint research team led by Professor Jeonghun Kwak of SNU and Professor Jaehoon Lim of Sunkyunkwan University has developed an AI-based platform that reverse-designs the optimal solvent properties needed to arrange quantum dots uniformly and densely during the fabrication of quantum dot light-emitting diodes. This research was supported by the Future Display Strategy Research Laboratory Support Program and the Nano and Material Technology Development Program, promoted by the Ministry of Science and ICT and the National Research Foundation of Korea. The findings were published online on July 15 in Reports on Progress in Physics, a leading international journal in physics published by the Institute of Physics (IOP) in the United Kingdom. QLEDs are devices that use quantum dots, nanometer-scale semiconductor particles, as their light-emitting layer. They are considered a promising technology for next-generation displays. To realize high-performance QLEDs, quantum dot particles must be arranged uniformly and densely within a thin film, much like bricks in a wall. The challenge is that, in this type of solution process, the brightness and lifetime of the device can vary significantly depending on which solvent is used to form the thin film. Because it is difficult to predict how specific solvent conditions affect device performance, researchers have largely relied on experience and repeated experiments to identify optimal conditions, resulting in significant time and cost limitations. To address this complexity, the research team trained AI to learn the relationship between the physical properties of solvents and the structure of quantum dot thin films. First, the team fabricated quantum dot thin films using five representative solvents and quantified how uniformly the surfaces were formed using atomic force microscopy (AFM).* They then trained a machine learning model on solvent properties, including vapor pressure, viscosity, density, and dielectric constant, as well as thin-film morphology data, enabling the model to inversely predict the solvent properties that could form the most uniform quantum dot thin film. Although no single solvent possessed all of the optimal properties proposed by the AI, the research team realized the AI-suggested conditions by combining multiple solvents. These were complex conditions that would have been difficult to identify through conventional repeated experiments alone. When applied to an actual QLED fabrication process, the optimized solvent system improved efficiency by approximately two times and operating lifetime by more than 40 times compared with conventional single-solvent processing. Professor Kwak stated, “This research demonstrates that AI can be used to design display materials and processes in a data-driven manner. We expect this approach to be applicable to the development of various next-generation electronic devices, including OLEDs and solar cells.” ▲ Figure 1. Conceptual diagram of the AI-based QLED process design platform (Left) In a conventional process, quantum dots are arranged unevenly, which interferes with charge transport and degrades device performance. (Right) The AI-based process design platform developed by the research team predicts the optimal solvent composition, enabling a more uniform arrangement of quantum dots. QLEDs fabricated using this platform achieved a twofold improvement in efficiency and a 40-fold improvement in lifetime. [Reference] - Paper/Journal: Machine-learning-enabled solvent engineering for uniform quantum dot packing in efficient and stable quantum-dot light-emitting diodes, Reports on Progress in Physics, 89, 078002 - DOI: https://doi.org/10.1088/1361-6633/ae8470 - Electronic Times (2026. 07. 15) : AI가 찾아낸 최적의 QLED 공정…효율 2배·수명 40배 향상 - DongA Science (2026. 07. 15) : AI로 공정 역설계…QLED 수명 40배 늘렸다 [Contact] Professor Jeonghun Kwak / Advanced Opto & Nano Electronics Laboratory / 02-880-1781 / jkwak@snu.ac.kr Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57895 Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr ...

SNU ECE research team led by Professor Jerald Yoo develops skin-adhesive wearable system that measures ECG signals without batteries - Proposes body-coupled wireless power supply technology that ensures human safety and stability - Addresses power supply challenge, a key obstacle to the commercialization of wearable devices ▲ (From left) Professor Jerald Yoo, Dr. Zhuoyue Li, and integrated M.S.-Ph.D. candidates Kyungsoo Park, Donghan Kim, and Gwangjin Kim Professor Jerald Yoo’s research team in the ECE Department announced that it has developed “SkinECG,” a skin-adhesive wearable healthcare system capable of measuring electrocardiogram signals without a battery. By combining energy harvesting technology with body-coupled power transfer, the research team proposed a new solution to one of the biggest challenges in the commercialization of wearable devices: power supply. The research was published on May 1 in Science Advances, an international academic journal published by the American Association for the Advancement of Science (AAAS). ■ Research Background Wearable healthcare systems are gaining attention as next-generation medical technologies that can measure biological signals in real time through sensors worn on the body and detect early signs of disease. A representative example is the electrocardiography (ECG) sensor. ECG sensors measure electrical signals generated by the heart and are essential for identifying cardiovascular diseases such as arrhythmia. However, batteries remain a major obstacle to the commercialization and long-term use of wearable devices. Due to their size and weight, batteries reduce wearability, and when they are discharged, biological signal collection may be interrupted. They also require periodic charging and replacement, causing inconvenience for users and making long-term continuous monitoring of biological signals difficult. To address this issue, previous studies have attempted to apply energy harvesting technology to wearable devices. Energy harvesting converts ambient energy, such as light, heat, and movement, into electricity. However, there has been a mismatch between the location where a wearable sensor must be attached and the location where energy can be efficiently harvested. For example, ECG sensors are generally attached to the chest, while power-generating devices such as solar cells are more efficient when attached to areas such as the arms or legs, where they can receive sunlight. In other words, the optimal location for generating electricity does not necessarily match the location where biological signals need to be measured. ■ Research Achievements To overcome this fundamental limitation, Professor Yoo’s research team proposed a new power supply architecture that wirelessly delivers power generated by multiple energy-harvesting devices attached to the human body to a remote ECG sensor. In the paper, this technology is referred to as an Orthogonal Energy Harvesting Network (O-EHN). SkinECG consists of an ECG sensor, which integrates a flexible circuit board and semiconductor chip on a skin-adhesive hydrocolloid patch, and a multi-energy wireless power supply network that delivers power generated by multiple energy-harvesting devices to the sensor. ▲ Figure 1. Conceptual diagram of battery-free wearable power supply technology The system operates by converting ambient energy into electricity through one or more energy-harvesting devices and then wirelessly supplying that power to an ECG sensor on the chest through body-coupled power transfer technology. Each power-generating device is designed to transmit power at an orthogonal frequency, allowing the number and placement of devices to be flexibly adjusted while ensuring stable power delivery to the ECG sensor. Professor Yoo’s team also overcame the limitations of conventional wireless power transfer methods. Conventional approaches transmit power by radiating electromagnetic waves through the air, but when used near the human body, electromagnetic waves may be absorbed or scattered by the body, reducing efficiency. The research team instead focused on transmitting power along the surface of the skin rather than radiating it over a distance. Using body-coupled powering, the team successfully delivered power generated by devices attached to the body to the ECG sensor on the skin without wires. The system was also designed so that power signals from multiple energy-harvesting devices do not interfere with one another by using distinct frequency channels, enabling stable power delivery to the sensor. In particular, the research team limited the level of power coupled to the human body to a level comparable to what people are routinely exposed to from surrounding electronic devices and everyday environments. The system was operated under low-power conditions designed with human safety in mind. Through this, the team demonstrated that an ECG sensor can be powered solely by energy harvesting, without batteries or wires. ■ Expected Impact The development of SkinECG is expected to mark an important turning point in solving the power supply challenge for next-generation wearable healthcare systems. This technology can also be applied not only to ECG monitoring, but also to long-term monitoring of various biological signals, including electromyography and electroencephalography. Furthermore, it is expected to develop into a foundational technology for addressing power supply challenges in wearable electronics and implantable medical devices. In addition, because the technology reduces constraints on the number and placement of energy-harvesting devices and can be combined with existing commercial energy harvesting technologies, it offers strong potential for expansion into a wide range of future wearable healthcare devices. ■ Researchers’ Remarks Professor Yoo explained, “Wearable healthcare devices have faced a fundamental limitation: the location where ambient energy can be effectively harvested and the location where biological signals must be measured are often different. This research addresses that problem by wirelessly delivering power along the surface of the human body.” He added, “We limited the power level delivered to the human body to a level comparable to everyday exposure, taking safety into account. Through this, we demonstrated that stable power can be supplied to ECG sensors without heavy and bulky batteries. In the future, this technology could be expanded not only into a multimodal digital healthcare platform that powers various biological signal sensors, such as electromyography and electroencephalography sensors, but also into a foundational power supply technology for a wide range of wearable devices.” ■ Researcher Career Path The first author of the paper, Dr. Zhuoyue Li, received her Ph.D. in February 2026 from the Department of Electrical and Computer Engineering at the National University of Singapore (NUS) under the co-supervision of Professor Jerald Yoo and worked as a visiting researcher at Seoul National University. Co-authors Kyungsoo Park, Donghan Kim, and Gwangjin Kim are conducting research in the field of body area networks (BAN). This study was conducted as an international joint research project led by Professor Jerald Yoo’s research team at Seoul National University, with participation from the University of Tokyo and the National University of Singapore. ▲ Figure 2. Implementation and demonstration of a battery-free skin-adhesive wearable system A solar cell-based wireless power supply module (left) and a skin-adhesive ECG sensor (center) were attached to the human body. The team successfully measured ECG signals while wirelessly supplying power without a battery (right). [Reference] - Paper/Journal : SkinECG: An orthogonal remote powering wearable skin-like sensor, Science Advances - DOI : https://doi.org/10.1126/sciadv.aec9803 - Chosun Ilbo (2026. 05. 20) : Wearable System Developed to Measure ECG Signals Without Batteries - Hankook Ilbo (2026. 05. 29) : Energy Harvesting, a Solution to the Charging Problem, Could Become Part of Everyday Life Within Five Years [Contact] Professor Jerald Yoo / High-Performance Integrated Microsystems Laboratory / 02-880-1776 / jerald@snu.ac.kr Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57886 Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr...

■ Research Background Photonic integrated circuits are gaining attention as a next-generation technology capable of processing information quickly and efficiently using light. In particular, in the fields of data centers, optical communications, and optical computing, technologies that go beyond simply transmitting optical signals at high speed are becoming increasingly important. These include synchronizing the arrival times of multiple signals and delaying signals when needed. To realize these functions, researchers have studied coupled-resonator-induced transparency (CRIT) structures, which use interference between multiple optical resonators. CRIT is an optical phenomenon that selectively transmits light within a specific frequency band and, in the process, can slow down the propagation speed of optical signals. However, conventional CRIT structures are largely fixed once fabricated, making it difficult to reconfigure the same circuit for different functions when application requirements change. For example, to delay optical signals for a longer period or shift them to a certain frequency band, a new optical device tailored to that specific function had to be designed. As such, optical communication equipment and data center systems have become more complex, while adding new functions has required significant time and cost. In environments such as AI servers and next-generation data centers, where massive amounts of data must be processed in real time, this lack of flexibility has been regarded as a major obstacle to the advancement of optical computing technologies. ■ Research Achievements To overcome this limitation, the joint research team proposed a new approach that treats the two optical states that constitute a CRIT system—the bright mode and the dark mode—as a unified system. The team also introduced two controllable loop couplers and established a new design principle for a programmable photonic integrated circuit that can reconfigure optical resonator structures, which were previously difficult to modify after fabrication, depending on specific requirements. The researchers came up with a new CRIT structure capable of delaying and controlling the flow of light and demonstrated that optical interference between the bright and dark modes can be treated as a single design variable. This significantly expands the design freedom of optical resonator circuits, whose structures had previously been fixed. In particular, the team proved that two loop couplers can be used to control the width and shape of the frequency band through which optical signals pass, as well as the delay and transmission characteristics of optical signals traveling through the circuit. This means that the propagation speed and transmission characteristics of optical signals can be freely reconfigured not only in a single optical resonator, but also across an entire structure composed of multiple connected resonators. The researchers also numerically demonstrated how the propagation speed of optical pulses changes in real time while the circuit is being actively controlled. As a result, they confirmed that the delay time of optical signals can be freely adjusted while largely maintaining signal-processing performance. They also verified that the frequency components of light can be converted without adding any special devices. Furthermore, through three-dimensional electromagnetic field simulations, the research team verified that the proposed CRIT device can be implemented on a silicon nitride (Si₃N₄) photonic integrated circuit platform. They also analyzed various factors that may arise during actual fabrication and operation, including material loss, variations in resonator quality, backscattering, changes in coupling characteristics, phase errors in loop couplers, and thermal crosstalk. The results confirmed that the proposed structure can operate stably even in realistic photonic integrated circuit environments. ■ Expected Impact This research is significant because it presents a new programmable photonic integrated circuit platform that goes beyond the limitations of conventional fixed optical signal delay structures and enables the temporal and frequency characteristics of optical signals to be controlled even while the circuit is operating. In particular, the study demonstrates the possibility of implementing key functions required for next-generation optical interconnects—such as optical signal synchronization, variable delay lines, optical buffers, and light-frequency conversion—within a single photonic integrated circuit structure. The photonic integrated circuit design method proposed by the research team can also be extended beyond CRIT to the dynamic control of various resonator-based optical circuits. This suggests that the design principle could serve as a foundation for next-generation optical signal processing technologies that design and control the flow of light according to specific needs. If the proposed photonic integrated circuit is commercialized in the future, it is expected to allow the speed of optical signals to be adjusted as needed while enabling a single optical chip to switch among various functions like software. As such, the technology is expected to reduce power consumption and improve data processing efficiency in data centers and AI servers. Additionally, because various signal-processing functions can be integrated into a single optical chip, the technology could contribute to the miniaturization and cost reduction of optical communication equipment and sensor systems. In the long term, it is expected to serve as a core enabling technology in a wide range of industries that require ultrafast information processing, including autonomous driving, next-generation communications, and quantum technologies. - Paper/Journal: Fully programmable slow light based on a spinor representation of generalized coupled-resonator-induced transparency, Advanced Science - DOI: https://doi.org/10.1002/advs.76378 Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57876 Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr...

Professor Jaeyoung Do’s research team at the AIDAS Lab in the Department of Electrical and Computer Engineering at Seoul National University announced that its paper, “VALUEFLOW,” which studies the alignment of large language models based on human values, has been accepted as an Oral Presentation at the International Conference on Machine Learning (ICML) 2026. ICML is one of the world’s most prestigious conferences in artificial intelligence and machine learning. At ICML 2026, which will be held in Seoul, Oral Presentation is a presentation format granted only to outstanding research, accounting for approximately 0.7% of all submitted papers. Figure 1. Architecture of VALUEFLOW Large language models are increasingly being used in a wide range of high-risk and high-value domains, including education, healthcare, policy, and decision-making support. As a result, technologies that allow models to understand and adjust the human values and judgment criteria underlying their responses are becoming increasingly important, beyond simply following user preferences. The Need for Human Value-Based AI Alignment Existing AI alignment research has primarily adjusted models based on user preferences or feedback scores. However, preferences can easily change depending on how a question is asked or the context in which it is presented, making it difficult to stably reflect the values that individuals or groups fundamentally consider important. For example, even for the same issue, some users may prioritize fairness, while others may place greater importance on freedom or safety. To address these limitations, the research team proposed VALUEFLOW, an integrated framework that can represent, measure, and steer the responses of large language models from the perspective of human values. VALUEFLOW: An Integrated Framework Connecting Value Representation, Measurement, and Steering VALUEFLOW consists of three main stages. First, in the Value Representation stage, the framework integrates different value theories into a single structured embedding space through HiVES, a hierarchical value embedding model. This enables the model to capture value signals at the text level, including care, fairness, freedom, safety, rights, and responsibility, as defined in value systems such as Schwartz’s theory of basic values, moral foundations theory, and rights- and duty-based value frameworks. Second, in the Value Measurement stage, VALUEFLOW uses the large-scale Value Intensity Database (VIDB) and a comparison-based evaluation method to quantitatively assess not only whether a particular value is present, but also how strongly it is expressed. Third, in the Value Steering stage, the framework guides model responses to reflect specific value directions and intensities, while analyzing the steerability and limitations of each model. Figure 2. Comparison of value steerability across major LLMs. The figure shows differences in model responses to positive and negative value steering. Analyzing the Value Steerability of 10 Major Large Language Models Using VALUEFLOW, the research team evaluated the steerability of 10 major large language models, including GPT-4.1, Gemini, Claude, Qwen, Mistral, Gemma, and Grok, across 4 value theories and 32 value categories. The analysis revealed clear differences in value-steering capabilities across models. Some models responded well to positively reinforcing certain values but showed little response to steering in a negative direction. In particular, values generally considered socially desirable, such as care and universalism, showed strong resistance to negative steering. The team also found that when multiple values were steered simultaneously, similar values tended to be reinforced together, while conflicting values tended to weaken each other’s expression. This suggests that value alignment in large language models is not merely a matter of following instructions, but is closely connected to the models’ internal safety and alignment characteristics. VALUEFLOW can be applied to personalized AI, culture-specific AI alignment, policy-sensitive AI deployment, model auditing, and real-time dialogue-based alignment. In particular, because it can analyze which values a model reflects well and which values it resists, VALUEFLOW is regarded as a foundational technology for transparent and responsible AI development. This research was led by Woojin Kim of the Department of Electrical and Computer Engineering at Seoul National University as the first author, with Sieun Hyeon and Jusang Oh as co-authors and Professor Jaeyoung Do as the corresponding author. The research team plans to expand VALUEFLOW to multi-turn dialogue-based personalized alignment, culture-specific value profiling, and multimodal AI alignment. Professor Do stated, “This research is meaningful because it presents a direction for large language models to move beyond simply following users’ immediate preferences and toward structurally understanding and steering the values that humans consider important. We will continue to develop this work into responsible AI alignment technology that enables the coexistence of diverse individual and societal values.” Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57850 Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr...

Won Hyo Jeong, a Ph.D. candidate in the Electric Energy Conversion Lab (EECL) led by Professor Jung-Ik Ha in the Department of Electrical and Computer Engineering, received the Outstanding Paper Award for Young Engineers at the 2026 International Power Electronics Conference, IPEC-Nagasaki 2026 -ECCE Asia-, held in Nagasaki, Japan, from May 31 to June 4. The award-winning paper, titled “Charge Pump Circuits for Negative Voltage Turn-Off in Gate Drivers,” presents a method for generating negative voltage for switch driving in high-voltage, high-current circuits using SiC (silicon carbide) switches without employing isolated DC-DC components. Held annually, ECCE Asia is one of Asia’s largest international conferences in the field of power electronics. The conference serves as a platform for sharing next-generation energy conversion technologies and the latest research achievements. Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57823 Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr...

Professor Jonghyun Choi’s research team in the Department of Electrical and Computer Engineering at Seoul National University has been selected for NVIDIA’s Academic Grant Program. The NVIDIA Academic Grant Program supports innovative artificial intelligence research at universities and accredited research institutions around the world. Selected through a review process, research teams receive access to the latest GPU infrastructure and research resources free of charge. As part of the program, Prof. Choi’s team will receive credits for approximately 34,000 hours of use on a Brev cloud platform node equipped with eight H100 80GB GPUs, as well as two RTX PRO 6000 Max-Q GPUs. Based on this support, the team plans to conduct research to improve the performance and generalization capabilities of Vision-Language-Action (VLA) models. VLA models are artificial intelligence models that understand visual information and language instructions and translate them into physical actions, effectively serving as the “brain” that enables robots to perform diverse tasks in various environments. Through this research, the team is expected to enhance the multi-task performance and generalization capabilities of VLA models and contribute to the development of core technologies in the field of Embodied AI. Source: https://ece.snu.ac.kr/ece/news?md=v&bbsidx=57808 Translated by: Changhoon Kang, English Editor of the Department of Electrical and Computer Engineering, changhoon27@snu.ac.kr...