Tianle Li
YOU?
Author Swipe
View article: Solar-Blind Mobile Deep Ultraviolet Optical Communication Utilizing Photomultiplier Tubes
Solar-Blind Mobile Deep Ultraviolet Optical Communication Utilizing Photomultiplier Tubes Open
Ozone in the atmosphere strongly absorbs deep ultraviolet light with wavelengths between 200 and 280 nm. Therefore, this characteristic is advantageous and promising for unperturbed, non-disturbed information transmission in fields such as…
View article: Stability Analysis Based on Hybrid αβ-impedance Model of Grid-Connected Inverters under Weak Grid
Stability Analysis Based on Hybrid αβ-impedance Model of Grid-Connected Inverters under Weak Grid Open
The robustness of the grid-connected inverter (GCI) system in weak grids is deteriorated due to consider discrete characteristics of the GCI control system. Under the same main circuit parameters and control loop parameters, the small sign…
View article: Reshaping Reasoning in LLMs: A Theoretical Analysis of RL Training Dynamics through Pattern Selection
Reshaping Reasoning in LLMs: A Theoretical Analysis of RL Training Dynamics through Pattern Selection Open
While reinforcement learning (RL) demonstrated remarkable success in enhancing the reasoning capabilities of language models, the training dynamics of RL in LLMs remain unclear. In this work, we provide an explanation of the RL training pr…
View article: Unveiling the Compositional Ability Gap in Vision-Language Reasoning Model
Unveiling the Compositional Ability Gap in Vision-Language Reasoning Model Open
While large language models (LLMs) demonstrate strong reasoning capabilities utilizing reinforcement learning (RL) with verifiable reward, whether large vision-language models (VLMs) can directly inherit such capabilities through similar p…
View article: Zeolite and silica fume as enhancers for cementitious solidification of borate radioactive waste
Zeolite and silica fume as enhancers for cementitious solidification of borate radioactive waste Open
Low- and medium-level borate radioactive liquid wastes are common liquid wastes generated during the operation and decommissioning of nuclear power plants. These wastes pose significant risks to environmental safety and public health, nece…
View article: Architecting heterostructures in multilayered titanium laminates to attain 1 GPa yield stress with uncompromised ductility at 500 °C
Architecting heterostructures in multilayered titanium laminates to attain 1 GPa yield stress with uncompromised ductility at 500 °C Open
Lightweight, high‐strength, and heat‐resistant protective structures have consistently been crucial for applications in extreme environments, such as aerospace, semiconductors, and nuclear power industries. Multilayered TC4/TB8 titanium (T…
View article: Oral colon-targeted delivery of recombinant human MANF for alleviation of ulcerative colitis
Oral colon-targeted delivery of recombinant human MANF for alleviation of ulcerative colitis Open
Midbrain astrocyte-derived neurotrophic factor (MANF) is a secreted protein induced by endoplasmic reticulum stress. Previous studies have indicated that intravenous administration of 1 mg/kg/day recombinant human MANF protein with His tag…
View article: Prompt-to-Leaderboard
Prompt-to-Leaderboard Open
Large language model (LLM) evaluations typically rely on aggregated metrics like accuracy or human preference, averaging across users and prompts. This averaging obscures user- and prompt-specific variations in model performance. To addres…
View article: Structure-Guided design of Cas12a variants improves detection of nucleic acids
Structure-Guided design of Cas12a variants improves detection of nucleic acids Open
CRISPR-Cas12a holds promising potential for pathogen detection. However, its performance is not optimal when combined with isothermal amplification. Hence, we engineered a mutant of LbCas12a (K595A) with reduced cis-cleavage activity, to m…
View article: Project MPG: towards a generalized performance benchmark for LLM capabilities
Project MPG: towards a generalized performance benchmark for LLM capabilities Open
There exists an extremely wide array of LLM benchmarking tasks, whereas oftentimes a single number is the most actionable for decision-making, especially by non-experts. No such aggregation schema exists that is not Elo-based, which could …
View article: How to Evaluate Reward Models for RLHF
How to Evaluate Reward Models for RLHF Open
We introduce a new benchmark for reward models that quantifies their ability to produce strong language models through RLHF (Reinforcement Learning from Human Feedback). The gold-standard approach is to run a full RLHF training pipeline an…
View article: The value of anxiety and depression in predicting physical function and major adverse cardiovascular events in patients with acute coronary syndrome
The value of anxiety and depression in predicting physical function and major adverse cardiovascular events in patients with acute coronary syndrome Open
Social support improved physical functionality and reduced the impact of psychological distress. Psychological state had the greatest long-term prognostic value in patients with CHD.
View article: The Parameter Calibration of Social Force Model for Pedestrian Flow Simulation Based on YOLOv5
The Parameter Calibration of Social Force Model for Pedestrian Flow Simulation Based on YOLOv5 Open
With the increasing importance of subways in urban public transportation systems, pedestrian flow simulation for supporting station management and risk analysis becomes more necessary. There is a need to calibrate the simulation model para…
View article: Global Optimization and Quantitative Assessment of Large-Scale Renewables-Based Hydrogen System Considering Various Transportation Modes and Multi-Field Hydrogen Loads
Global Optimization and Quantitative Assessment of Large-Scale Renewables-Based Hydrogen System Considering Various Transportation Modes and Multi-Field Hydrogen Loads Open
In the past, hydrogen was mostly produced from fossil fuels, causing a certain degree of energy and environmental problems. With the development of low-carbon energy systems, renewable energy hydrogen production technology has developed ra…
View article: From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipeline
From Crowdsourced Data to High-Quality Benchmarks: Arena-Hard and BenchBuilder Pipeline Open
The rapid evolution of Large Language Models (LLMs) has outpaced the development of model evaluation, highlighting the need for continuous curation of new, challenging benchmarks. However, manual curation of high-quality, human-aligned ben…
View article: GenAI Arena: An Open Evaluation Platform for Generative Models
GenAI Arena: An Open Evaluation Platform for Generative Models Open
Generative AI has made remarkable strides to revolutionize fields such as image and video generation. These advancements are driven by innovative algorithms, architecture, and data. However, the rapid proliferation of generative models has…
View article: MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark
MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark Open
In the age of large-scale language models, benchmarks like the Massive Multitask Language Understanding (MMLU) have been pivotal in pushing the boundaries of what AI can achieve in language comprehension and reasoning across diverse domain…
View article: Multi-parameter fusion diagnosis for medium and lower voltage switchgear cabinet based on UHF and Infrared Camera Method
Multi-parameter fusion diagnosis for medium and lower voltage switchgear cabinet based on UHF and Infrared Camera Method Open
Multi-parameter live detection and fusion diagnosis of medium and lower voltage switch-gear cabinet (M-LVSC) is an important technology for identification the operating status of power equipment. Under the operating conditions, a single se…
View article: Long-context LLMs Struggle with Long In-context Learning
Long-context LLMs Struggle with Long In-context Learning Open
Large Language Models (LLMs) have made significant strides in handling long sequences. Some models like Gemini could even to be capable of dealing with millions of tokens. However, their performance evaluation has largely been confined to …
View article: Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference
Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference Open
Large Language Models (LLMs) have unlocked new capabilities and applications; however, evaluating the alignment with human preferences still poses significant challenges. To address this issue, we introduce Chatbot Arena, an open platform …
View article: DTFFS: Dimension Transformation for Food Safety Risk Assessment Method
DTFFS: Dimension Transformation for Food Safety Risk Assessment Method Open
As people’s lives become more fast-paced, many unscrupulous businesses adulterate food during the supply process, resulting in serious food safety incidents. Therefore, food safety assessment is essential to the food supply chain. Research…
View article: The crucial regulatory role of type I interferon in inflammatory diseases
The crucial regulatory role of type I interferon in inflammatory diseases Open
Type I interferon (IFN-I) plays crucial roles in the regulation of inflammation and it is associated with various inflammatory diseases including systemic lupus erythematosus (SLE), rheumatoid arthritis (RA), and periodontitis, impacting p…
View article: Proteomic and single-cell analysis shed new light on the anti-inflammatory role of interferonβ in chronic periodontitis
Proteomic and single-cell analysis shed new light on the anti-inflammatory role of interferonβ in chronic periodontitis Open
Periodontitis, a condition that results in periodontal attachment loss and alveolar bone resorption, contributes to the global burden of oral disease. The underlying mechanism of periodontitis involves the dysbiosis and dyshomeostasis betw…