Detector Technology (United States)

Reward-model-based fine-tuning is a central paradigm in aligning Large Language Models with human preferences. However, such approaches critically rely on the assumption that proxy reward models accurately reflect intended supervision, a condition often violated due to annotation noise, bias, or limited coverage. This misalignment can lead to undesirable behaviors, where models optimize for flawed signals rather than true human values. In this paper, we investigate a novel framework to identify and mitigate such misalignment by treating the fine-tuning process as a form of knowledge integration. We focus on detecting instances of proxy-policy conflicts, cases where the base model strongly disagrees with the proxy. We argue that such conflicts often signify areas of shared ignorance, where neither the policy nor the reward model possesses sufficient knowledge, making them especially susceptible to misalignment. To this end, we propose two complementary metrics for identifying these conflicts: a localized Proxy-Policy Alignment Conflict Score (PACS) and a global Kendall-Tau Distance measure. Building on this insight, we design an algorithm named Selective Human-in-the-loop Feedback via Conflict-Aware Sampling (SHF-CAS) that targets high-conflict QA pairs for additional feedback, refining both the reward model and policy efficiently. Experiments on two alignment tasks demonstrate that our approach enhances general alignment performance, even when trained with a biased proxy reward. Our work provides a new lens for interpreting alignment failures and offers a principled pathway for targeted refinement in LLM training.

Metasurface holography has emerged as a transformative technology with significant potential in information encryption, virtual displays, and high-capacity data storage. However, developing a holographic display system with ultrahigh information capacity and multidimensional multiplexing remains a major challenge. To address this issue, we propose a Janus metasurface holographic display scheme featuring asymmetric transmission and multi-depth imaging at microwave frequencies, enabled by complex-amplitude modulation. The modified three-dimensional holography method we proposed enables the complex-amplitude modulation devices to acquire the ability of manipulating multi-depth holographic displays. By simultaneously multiplexing the propagation direction and propagation distance of incident waves, a large number of distinct holographic images can be reconstructed throughout the entire propagation space. As a proof of concept, we experimentally demonstrate the reconstruction of five distinct letter images, each located on different planes along opposite propagation directions. The measured results show excellent agreement with simulations, validating the proposed strategy. This approach paves the way for metasurface-based holographic display systems featuring ultrahigh information capacity, robust holographic encryption, and high-efficiency data storage.

Multi-Object Tracking (MOT) aims to build moving trajectories of objects within video sequences and serves as a critical component in autonomous driving systems. Recently, several studies have revealed the vulnerability of existing MOT methods by investigating adversarial attacks against MOT, raising significant safety concerns for real-world applications. These methods attack trackers by deliberately inserting false alarms, which mislead trajectories to drift from their correct paths. However, current MOT attack methods fail to propose efficient strategies for generating false alarms, as they either rely on computationally intensive optimization to determine the placement of false alarms, or crudely insert a large number of heuristically designed false alarms. In this paper, we propose an explainable and effective false alarm generation module, named Target Generating Module (TGM), that adaptively determines the location and size of false alarms by leveraging historical trajectory information. Based on this module, we design an attack method targeting mainstream MOT approaches, named <bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">T</b>rajectory-<bold xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</b>ware Attack (TA Attack). TA Attack achieves effective disruption of MOT systems by combining detection erasure and false alarm generation, requiring only a few frames to successfully compromise trajectories. To exhibit the flexibility and effectiveness of our method, we conduct experiments using four multi-object trackers (ByteTrack, SORT, CenterTrack and FairMOT) which are enabled by two representative detectors (YOLOX and CenterNet). The results demonstrate our method achieves state of the art performance with 74.87% attack success rate on BDD100K, 81.7% attack success rate on MOT17 and 83.87% attack success rate on MOT20 while 4 frames being attacked averagely, revealing the vulnerability of association mechanism in MOT methods.

Condensation flow of FC-72 in a rectangular microchannel made of polydimethylsiloxane with soft wall deformation was studied. The microchannel has a rectangular cross-sectional shape with hydraulic diameters of 0.222, 0.240, and 0.286 mm. Mass fluxes of 150–380 kg/(m2‧s) and vapor mass qualities of 0.2–0.6 were experimentally tested. The images of the two-phase flow pattern were captured using a high-speed camera. The deformation of the elastic channel's upper wall was measured using a white light confocal coaxial displacement sensor. The upper wall deformation, represented by an increase in thickness, occurred during intermittent flow. The maximum variation in the upper wall thickness of the channel is up to 25% of the original wall thickness, which was increased with increasing inlet vapor mass quality and mass flux. Surprisingly, the channel wall tended to thicken when sequential bubbles flow through. This phenomenon came from the local fluid pressure decrease caused by condensation, which was strong enough to make the bubble shrink rapidly. This shrinking may leave a temporary vacuum in the liquid slug connecting two adjacent bubbles. In channels with soft wall, the vacuum pulls the flexible wall moving toward the inside of the channel, resulting in wall thickening observed in the experiments. Therefore, this wall deformation is caused by condensation. A theoretical model based on phase change flow and elasticity equations was proposed, whose results agreed well with the experimentally measured deformation. The outcomes can serve as a heat dissipation solution for flexible devices in high-tech fields such as medical and aerospace applications.

This paper presents the design of a cryogenic operational transconductance amplifier (OTA) operating in the liquid nitrogen temperature (77K), targeting the low-temperature operational requirements of infrared imaging sensors. Guided by established 77K low-temperature models, a dual-mode (room temperature/cryogenic) OTA was fabricated and tested. This work details the modeling process, simulation results, and experimental data, validating the performance enhancement potential of low-temperature operation for circuits.

<strong class="journal-contentHeaderColor">Abstract.</strong> Accurate retrieval of cloud optical and microphysical properties (COMP) at night is important for monitoring changes in weather and climate systems. The nighttime cloud optical and microphysical properties (NCOMP) retrieval is enhanced by integrating data from hyperspectral infrared sounder and high-resolution imager on the same geostationary platform with a machine learning framework. Using geostationary satellite imager broadband thermal infrared (TIR) channels along with dozens of optimally selected hyperspectral IR (HIR) channels, we demonstrate substantial improvements over traditional TIR-channel-based methods. The HIR channels enhance sensitivity to cloud effective radius (CER) and optical thickness (COT), particularly for optically thin clouds, reducing retrieval errors to 9.73 &mu;m and 6.09, respectively, with an approximate 10 % accuracy improvement. The ML-based model preserves strong day-night continuity in COMP retrievals and assures the diurnal information for clouds, although challenges remain for thick clouds. This work highlights the importance of GEO-satellite-based HIR sounders, which provide critical spectral information that complements imager data for cloud optical and microphysical property retrievals. Middle-wave IR (MWIR) channels significantly improve COT retrieval. The proposed fusion approach offers a flexible retrieval framework applicable to future geostationary satellite systems for enhancing the cloud property retrievals containing diurnal information.