Publications

Recent video generators can fabricate realistic depictions of wars, disasters, public emergencies, and other real-world crises, creating substantial risks of misinformation. Existing benchmarks, however, provide limited evidence on detector and generator behavior in such settings, including how detectability varies with generation conditions, how people perceive generated videos, and whether detectors remain reliable during social dissemination. To address this gap, we introduce RA-Bench, a benchmark for AI-generated video detection that uses real videos as anchors. RA-Bench contains 17,886 videos, comprising 1,830 real-video anchors across 10 social-risk categories and 16,056 generated clips from four open-source and five closed-source generators. Based on RA-Bench, we organize our evaluation along three dimensions. We first assess detector generalization across seven traditional detectors, ten zero-shot multimodal models under three review settings, and two multimodal large language models specifically fine-tuned on AI-generated video detection. Across these methods, none of the three detector families generalizes consistently across RA-Bench instances. We then examine how detectability varies with generation quality, conditioning information, and sampling seeds. These analyses show that generation properties affect detector families differently, while source-level detection patterns remain stable across seeds. Finally, we study human authenticity judgments and detector reliability during social dissemination. We find that videos that mislead people are also difficult for current detectors, and that social dissemination makes detection harder. Together, these findings show that current methods struggle to detect realistic AI-generated videos, highlighting the need for detectors robust to evolving video generators.

High-performance computing underpins modern artificial intelligence, enabling foundation models, real-time inference and perception in autonomous systems, and data-intensive scientific simulations. Recent advances in quantization techniques using low-precision computation without degrading model accuracy create new opportunities for analog photonic computing characterized by ultra-high clock rates and low energy consumption. Here we propose and demonstrate a coherent homodyne integrated circuit capable of general matrix multiplication with aggregate throughput exceeding 1,000 tera-operations per second, enabled by massive on-chip optical fanout and parallelism. By leveraging time multiplexing, the required modulator count is reduced from quadratic to linear scaling, allowing dense integration of record-scale 256 by 256 homodyne units within a single reticle. The system achieves up to 7-bit computational accuracy across 8 by 8 parallel channels at a record computing clock rate of 120 Gbaud/s, and 6-bit statistical accuracy across 256 by 100 channels at 20–128 Gbaud/s, representing total throughput of 1,000–6,000 TOPS. Massive parallelism amortizes optoelectronic conversion to allow 330 TOPS/W efficiency using foundry-available packaging technology. The system throughput is benchmarked with Qwen2.5-0.5B models that generate accurate tokens. High throughput and energy efficiency establish a near-term pathway toward light-based accelerators for large-scale training and low-latency inference from data centers to edge devices.

Modern problems in high-performance computing, ranging from training and inferencing deep learning models in computer vision and language models to simulating complex physical systems with nonlinearly-coupled equations, require exponential growth of computational resources. Photonic analog systems are emerging with solutions of intrinsic parallelism, high bandwidth, and low propagation loss. However, their application has been hindered by low analog accuracy due to electro-optic distortion, material nonlinearities, and signal-to-noise ratios. Here we overcome this barrier with a quantization-aware digital-photonic mixed-precision framework across chiplets for accelerated AI processing and physical simulation. Using lithium niobate photonics with channel equalization techniques, we demonstrate linear multiplication with 9-bit amplitude-phase decoupling in homodyne optical logic with 6-bit precision at a clock rate of 128 giga-symbols per second, enabling AI processing with 6 ns latency. Co-designed hardware and algorithms, including iterative solvers, sparse-dense quantization, and bit-sliced matrix multiplication, explore photonic amplitude and phase coherence for complex-valued, physics-inspired computation. In electromagnetic problems, our approach yields 12-bit solutions for partial differential equations in scattering problems that would conventionally require up to 32-bit and often 64-bit precision. These results preserve digital-level fidelity while leveraging high-speed, low-energy photonic hardware, establishing a pathway toward general-purpose optical acceleration for generative artificial intelligence, real-time robotics, and accurate simulation for climate challenges and biological discoveries.

Large language models (LLMs) are increasingly adapted into domain-specific variants for applications in law, healthcare, and finance. Their scale, however, limits deployment in resource-constrained settings, and existing compression approaches often either degrade after domain adaptation or require substantial additional computation. We introduce EfficientXpert, a lightweight framework for domain pruning that integrates ForeSight Mask, a propagation-aware criterion for selecting weights to prune without backpropagation, and Partial Brain Surgeon, an efficient closed-form update for low-rank adapters under a fixed sparsity pattern. With fine-tuning cost comparable to standard LoRA, EfficientXpert converts a general pretrained model into a sparse, domain-adapted expert in a single pruning step. Across health and legal benchmarks, EfficientXpert reaches up to 98 percent of dense performance at 40 percent sparsity, improving over prior pruning baselines while matching LoRA training time and staying within 1 percent of LoRA peak GPU memory in our experiments.