Tsinghua University

Heterogeneous Treatment Effect Estimation with Distribution Shift and Unmeasured Confounding

Heterogeneous treatment effect (HTE) estimation is vital for understanding how treatment effects vary across individuals or subgroups. Recently, numerous HTE models, such as balancing methods and instrumental variable (IV) methods, have been developed to eliminate confounding bia...

Liquid Metal Multi‐Functional Droplets Enabled High‐Performance Tumor Therapy With Four Synergistic Mechano‐Thermal‐Chemo‐Freezing Effects

ABSTRACT Cryoablation, which induces rapid intracellular ice crystal formation, has been widely used in clinics for treating various tumors, yet monotherapy yields limited efficacy due to inadequate elimination of tumor cells. Here, we proposed an effective synergistic strategy w...

Virtual recreation of urban heritage of Jingdezhen for immersive VR display

To unify the decoupled dimensions of scientific rigor and contextual narrative in virtual heritage display, this paper uses the VR project developed for Jingdezhen’s bid for World Cultural Heritage as a case study, and establishes a workflow for the virtual recreation of urban he...

In situ reduction-driven formation of sub-nanometer alternating heterostructures

Sub-nanometer alternating heterostructures provide a platform for modulating electronic structure and interfacial reactivity through strong atom-level coupling, with potential applications in energy conversion. However, their precise construction remains challenging. Here we show...

Sharp Rank Transitions and Supercritical Kernel Structure in Sparse Cayley Incidence Matrices

Fix k ≥ 3 and let p → ∞ through odd primes. We sample m = ⌊Cp⌋ distinct rows uniformly from the additive Cayley host on kp part-labeled vertices and form their binary incidence matrix. Although local peeling gives a scalar prediction, the additive constraint leaves global row dep...

Trajectory-Constrained Feature Transformation Optimization Framework for Stable Tabular Data Reprogramming

Tabular data reprogramming is a cornerstone of Data-Centric AI, which seeks to mathematically reconfigure raw features into more expressive and model-efficient representations. This process has traditionally relied on extensive manual engineering. Further, recent automated featur...

Numerical Investigation of Suspension Heating for Solid Particles

With the global energy transition and large-scale renewable integration, thermal power units are shifting from base-load to peak-shaving roles, but deep peak shaving may cause combustion instability and increased emissions. High-temperature solid-particle thermal storage offers a...

Enhancing Knowledge Management in Green Building Engineering: A Hybrid Retrieval-Augmented Generation Approach to Support Project Assessment

Abstract Green building assessment relies heavily on manual queries of complex engineering regulations, making the introduction of knowledge management theories crucial for improving assessment efficiency. However, the inherent characteristics of engineering regulations—dense ter...

Balancing emission reduction and regulatory burden under inspection limits for diesel fleets

Effectively regulating in-use vehicle emissions requires methods that can identify high-impact emitters within heterogeneous vehicle fleets. This study develops a vehicle-level NOx accounting method based on inspection data, converting measured NOx concentrations into estimated e...

Receptive-field remapping and trans-saccadic space updating

Since its discovery, perisaccadic receptive-field (RF) remapping—the shift of RFs around the time of saccades—has inspired many studies exploring its origins, characteristics, and functional implications. We review recent developments leading to a circuit model that links spatiot...

Cornus mas L. Ameliorates Hepatic Insulin Resistance in High-fat Diet-induced Diabetic Mice

Background Type 2 diabetes (T2D) remains a prominent chronic metabolic disease characterized by hyperglycemia worldwide. Insulin resistance is the primary risk factor contributing to T2D development. Cornus mas L. is a natural product shown to play a regulatory role in metabolic...

Constructing a Xiangyunsha Defect Dataset and a Lightweight Detection Model for Sustainability

Xiangyunsha (Gambiered Guangdong silk) is a naturally dyed, handcrafted textile whose sustainable value is constrained by material waste arising from conventional defect-rejection practices. This study reconceptualizes conventionally defined defects, particularly crackle irregula...

Uncovering the hidden resource cost of urbanization: A globally scalable open-source framework for urban housing vacancy monitoring

Rapid urbanization often outpaces population growth, creating massive resource inefficiencies. Quantifying this structural mismatch between housing supply and actual occupancy remains a global challenge due to costly censuses and imprecise proxy data. Here, we present a globally...

Worth Their Weight: Randomized and Regularized Block Kaczmarz Algorithms without Preprocessing

Abstract. Due to the ever growing amounts of data leveraged for machine learning and scientific computing, it is increasingly important to develop algorithms that sample only a small portion of the data at a time. In the case of linear least-squares, the randomized block Kaczmarz...

Uniform and Robust NiO x /SAM Layer in Perovskite Solar Cell via Multifunctional Crown Ether Modification

The combination of self‐assembled molecules (SAMs) and nickel oxide (NiO x ) has recently become one of the basic structures for achieving high‐performance inverted perovskite solar cells (PSCs). However, the dissolution of SAM and other modification additives caused by subsequen...

More than regulation: How policy mixes shape innovation ecosystems in digital transformation

The era of digital transformation drives the continuous restructuring of innovation ecosystems, challenging regulators to move beyond conventional regulations to address market and system failures. Although policy mixes are essential to address these challenges and promote innova...

Infrared spectroscopy of MO(CH4) n (M = Ce, Y; n = 1–3) reveals two distinct adsorption configurations

Investigation of the interactions of methane with metal oxides is crucial for elucidating absorption configurations at active sites of catalysts, which has proven to be a challenging experimental target due to the difficulty in size selection of neutral systems. In this work, the...

Recognition and mechanism of M. tuberculosis cell wall galactan transport

Mycobacterial-derived lipid-linked galactan (LLG) is an essential component of the Mycobacterial tuberculosis (Mtb) cell wall core arabinogalactan (AG), a well-established virulence factor and target of anti-tubercular drugs. Cross-membrane translocation of LLG is energetically u...

Bias-reconfigurable barrier photodetectors for integrated broadband sensing and neuromorphic vision

Intelligent vision systems demand photodetectors with reconfigurable response to support diverse modalities like high-speed sensing and neuromorphic processing. Here we show a bias-reconfigurable barrier photodetector that enables continuous modulation of response time across 7 o...

A diffusion-based sequential generative workflow for AI-assisted architectural and structural design of detached houses

Traditional architectural and structural design workflows often require repetitive manual operations. Although recent artificial intelligence methods have advanced floor plan generation and structural design, most studies focus on isolated subtasks and simple residential datasets...