Publications (21)
Scaling Speculative Decoding with Lookahead Reasoning
Yichao Fu, Rui Ge, Zelei Shao +2
Reasoning models excel by generating long chain-of-thoughts, but decoding the resulting thousands of tokens is slow. Token-level speculative decoding (SD) helps, but its benefit is…
Leveraging Unstructured Text Data for Federated Instruction Tuning of Large Language Models
Rui Ye, Rui Ge, Yuchi Fengting +3
Federated instruction tuning enables multiple clients to collaboratively fine-tune a shared large language model (LLM) that can follow humans' instructions without directly sharing…
On-chip Lithium Niobate Heterogeneous Photonic Crystal Nanocavity Laser
Xiangmin Liu, Rui Ge, Chengyu Chen +3
Thin film lithium niobate (TFLN) has become an platform for modern integrated circuits due to its excellent optical properties. With the development of rare earth ion doped TFLN, i…
High Q and high gradient performance of the first medium-temperature baking 1.3 GHz cryomodule
Jiyuan Zhai, Weimin Pan, Feisi He +36
World's first 1.3 GHz cryomodule containing eight 9-cell superconducting radio-frequency (RF) cavities treated by medium-temperature furnace baking (mid-T bake) was developed, asse…
Simultaneous - and - nonlinear processes generation in thin film lithium tantalate microcavity
Xiongshuo Yan, Miao Xue, Jiangwei Wu +4
On-chip efficient nonlinear functions are instrumental in escalating the utilities and performance of photonic integrated circuits (PICs), especially for a wide range of classical…
FedLLM-Bench: Realistic Benchmarks for Federated Learning of Large Language Models
Rui Ye, Rui Ge, Xinyu Zhu +5
Federated learning has enabled multiple parties to collaboratively train large language models without directly sharing their data (FedLLM). Following this training paradigm, the c…
Chip-scale Spontaneous Quasi-Phase-Matched Micro-Racetrack Resonator
Tingge Yuan, Yi'an Liu, Xiongshuo Yan +6
Due to their capacity for non-classical light generation, high-efficiency second-order nonlinear parametric processes play an important role in quantum photonic technology, and chi…
Frequency tunable topological edge states of two-dimensional honeycomb lattice photonic crystals
Yuchen Peng, Bei Yan, Jianlan Xie +5
In this paper, the photonic quantum spin Hall effect (PQSHE) is realized in dielectric two-dimensional (2D) honeycomb lattice photonic crystal (PC) by stretching and shrinking the…
A Formal Model to Prove Instantiation Termination for E-matching-Based Axiomatisations (Extended Version)
Rui Ge, Ronald Garcia, Alexander J. Summers
SMT-based program analysis and verification often involve reasoning about program features that have been specified using quantifiers; incorporating quantifiers into SMT-based reas…
Mid-infrared single photon detector with superconductor MoSi nanowire
Qi Chen, Rui Ge, Labao Zhang +11
A mid-infrared single photon detector (MIR-SNSPD) was reported based on 30 nm-wide superconductor molybdenum silicide nanowires in this work. Saturated quantum efficiencies (QEs) w…
A 16-pixel NbN nanowire single photon detector coupled with 300 micrometer fiber
Qi Chen, Biao Zhang, Labao Zhang +8
Niobium Nitride (NbN) nanowire is the most popular detection material of superconducting nanowire single photon detectors (SNSPDs) for high repetition rate and high efficiency. How…
Commissioning of te China-ADS injector-I testing facility
Fang Yan, Huiping Geng, Cai Meng +58
The 10 MeV accelerator-driven subcritical system (ADS) Injector-I test stand at Institute of High Energy Physics (IHEP) is a testing facility dedicated to demonstrate one of the tw…
SUPERChem: A Multimodal Reasoning Benchmark in Chemistry
Zehua Zhao, Zhixian Huang, Junren Li +28
Current benchmarks for evaluating the chemical reasoning capabilities of Large Language Models (LLMs) are limited by oversimplified tasks, lack of process-level evaluation, and mis…
MAS-GPT: Training LLMs to Build LLM-based Multi-Agent Systems
Rui Ye, Shuo Tang, Rui Ge +4
LLM-based multi-agent systems (MAS) have shown significant potential in tackling diverse tasks. However, to design effective MAS, existing approaches heavily rely on manual configu…
Large Quality Factor Enhancement Based on Cascaded Uniform Lithium Niobate Bichromatic Photonic Crystal Cavities
Rui Ge, Xiongshuo Yan, Zhaokang Liang +5
In this paper, by cascading several bichromatic photonic crystals we demonstrate that the quality factor can be much larger compared with that in an isolated cavity without increas…
Coupled cavity-waveguide system based on topological corner state and edge state
Aoqian Shi, Bei Yan, Rui Ge +5
Topological corner state (TCS) and topological edge state (TES) have provided new approaches to control the propagation of light. The construction of topological coupled cavity-wav…
Medium-temperature furnace bake of Superconducting Radio-Frequency cavities at IHEP
Feisi He, Weimin Pan, Peng Sha +16
Recently, heat treatment between 250 C and 500 C has been attempted to improve quality factor of superconducting radio-frequency cavities at FNAL and KEK. Experiments of such mediu…
Doubly resonant photonic crystal cavity using merged bound states in the continuum
Rui Ge, Xiangmin Liu, Xiongshuo Yan +2
In this work, a doubly resonant photonic crystal (PhC) cavity using the merged bound states in the continuum (BICs) is proposed to obtain a higher second harmonic generation (SHG)…
The Tool Decathlon: Benchmarking Language Agents for Diverse, Realistic, and Long-Horizon Task Execution
Junlong Li, Wenshuo Zhao, Jian Zhao +18
Real-world language agents must handle complex, multi-step workflows across diverse Apps. For instance, an agent may manage emails by coordinating with calendars and file systems,…
High-Order Nonreciprocal Add-Drop Filter
Hang Li, Rui Ge, Yuchen Peng +4
Topological photonics have led to the robust optical behavior of the device, which has solved the problem of the influence of manufacturing defects and perturbations on the device…
Internalizing Agency from Reflective Experience
Rui Ge, Yichao Fu, Yuyang Qian +4
Large language models are increasingly deployed as autonomous agents that must plan, act, and recover from mistakes through long-horizon interaction with environments that provide…