Shanghai Jiao Tong University

Seeing Is Coding: On the Effectiveness of Vision Language Models in Code Understanding

Large Language Models (LLMs) have achieved remarkable success in source code understanding, yet as software systems grow in scale, computational efficiency has become a critical bottleneck. Currently, these models rely on a text-based paradigm that treats source code as a linear...

XSearch: Explainable Code Search via Concept-to-Code Alignment

With the emergence of deep learning, semantic code search has been widely adopted in both academia and industry. These approaches embed natural-language queries and code snippets into a shared embedding space and retrieve results based on vector similarity. Despite their strong p...

Pola-R-CHP in previously untreated DLBCL with international prognostic index score 0–1: a multicenter real-world retrospective study in China

Abstract Purpose Approximately 30% of newly diagnosed diffuse large B-cell lymphoma (DLBCL) cases are categorized as low-risk by the international prognostic index (IPI). Polatuzumab vedotin, rituximab, cyclophosphamide, doxorubicin, and prednisone (Pola-R-CHP) has demonstrated e...

Compiling Large Multi-modal Requirement Documents into Runnable Software Systems: From an Agentic Test-Driven Perspective

Large Language Models (LLMs) have significantly improved programming efficiency by parsing natural language into code snippets. However, their performance degrades significantly as requirements scale; when faced with multi-modal documents containing hundreds of scenarios, LLMs of...

IssueExec: A Test-Driven Approach for Localizing Software Engineering Issues

Issue localization, which identifies code locations requiring modification from issue descriptions, is a critical step in automated software maintenance. Existing approaches predominantly attempt to directly align issue descriptions with code elements, yet often struggle due to t...

Surface Patterns for Optical Modulation

Abstract Surface patterns with micro- and nanoscale features have emerged as a transformative paradigm in optics due to their ability to extend beyond the functional limits of conventional optical elements through strong light–matter interactions. By engineering geometric paramet...

Staged Multi-step UTXO Workflows via Recursive Invariants

Stateless UTXO-style execution validates transactions using local and referenced data, enabling parallel validation and predictable serialized-size/weight accounting. However, multi-step workflows must thread state across outputs, and a prepared next-step transaction may become s...

Digital communication and narrative governance in BRICS healthcare: a website media analysis of public hospital vision and mission statements

Abstract: This study conducts a qualitative thematic analysis of vision and mission statements from leading publicly listed hospitals in BRICS countries-Brazil, Russia, India, China, and South Africa-to explore institutional health priorities and strategic orientations. Drawing o...

Efficient Grammar-Constrained Decoding via Parser Stack Classification

LLMs are widely used to generate structured output like source code or JSON. Grammar-constrained decoding (GCD) can guarantee the syntactic validity of the generated output, by masking out tokens that violate rules specified by a context-free grammar. However, the online computat...

Generating Project-Specific Test Cases with Requirement Validation Intention

Test cases are valuable assets for maintaining software quality. State-of-the-art automated test generation techniques typically focus on maximizing program branch coverage or translating focal methods into test code. However, in contrast to branch coverage or code-to-test transl...

SINGLE-PHANTOM CT-REFERENCED EVALUATION OF RESIDUAL MEASUREMENT BIAS IN FULL-LENGTH LOWER-LIMB RADIOGRAPHY AFTER SID/SOD CALIBRATION

Purpose: This single-phantom CT-referenced technical-validation study evaluated residual length and angular measurement bias after source-to-image distance/source-to-object distance (SID/SOD) calibration in full-length lower-limb radiography using system/protocol configurations r...

Realization of Quantum Spin Hall Insulator Superlattice with Emergent Multigap‐Like Helical Edge States

ABSTRACT The functional quantum spin Hall insulators (QSHI), protected by time‐reversal symmetry against single‐particle backscattering, hold great promise for dissipationless quantum electronics. Realization of QSHI with gapped helical edge states, which would enable determinist...

Microfluidic-based high-throughput isolation enhances the recovery of novel strains and diversity from Arctic soil microbiome

ABSTRACT Microbial cultivation remains essential for understanding the physiology, ecology, and biotechnological potential of environmental microbes, yet conventional plate-based methods (CPM) recover only a minute fraction of the environmental microbiome. Polar regions, particul...

The ethics of influence: organizational culture as a moderator of leadership style in agile projects

Abstract: This research investigates the moderating role of organizational culture (OC) in shaping the relationship between leadership style and project performance (PP), with a specific focus on despotic and transformational leadership. Through a dual study model the data was co...

Derivative-Informed Graph Convolutional Autoencoder with Phase Classification for the Lifshitz-Petrich Model

The Lifshitz-Petrich (LP) model is a classical model for describing complex spatial patterns such as quasicrystals and multiphase structures. Solving and classifying the solutions of the LP model is challenging due to the presence of high-order gradient terms and the long-range o...

A novel dynamic model parameters identification method for robotic manipulator based on the feasible physical constraints

The dynamic model of robots plays a crucial role in control tasks. The traditional Coulomb-viscous friction model inadequately describes friction torque, and the conventional identification strategy is ineffective in accurately identifying dynamic models with full physical parame...

ANGPTL4 is a hypoxia associated candidate linked to immune and vascular phenotypes in ccRCC

Clear cell renal cell carcinoma (ccRCC) is characterized by prominent vascularization and cellular heterogeneity, but reliable biomarkers for angiogenesis and response to targeted therapy remain limited. In this study, transcriptomic data from GEO cohorts (GSE53757, GSE66271, GSE...

Analysis and prediction of research vessel attitude based on full-scale operational data

This study uses full-scale operational data from a large research vessel to perform statistical characterization and guide the selection of data-driven prediction models. A simplified Long Short-Term Memory-based (LSTM-based) framework, referred to as LSTM-FC, is developed for co...

DentalGEN: a large-scale controllable generative AI framework for automated dental crown restoration

Dental crown restoration addresses a global public health burden, yet is constrained by a slow, variable, and labor-intensive manual design process that limits access to high-quality treatment. Here we introduce DentalGEN, a large-scale controllable generative artificial intellig...

EaMIC3 inhibits cell apoptosis and facilitates Eimeria acervulina infection through interacting with EpCAM

Coccidiosis, caused by Eimeria acervulina, is a pathogenic protozoan disease, leading to huge economic losses to the global poultry industry. At the early stage of E. acervulina infection, microneme protein 3 (EaMIC3) is involved in the inhibition of infected host cell apoptosis,...