arXiv subjects

Zhu Liang

Publications and source records attributed to Zhu Liang.

6 recordsLinked to original sources

arXiv2026

Bioinfoysis Technical Report

Large language model agents have shown promise in bioinformatics, but most existing systems focus primarily on producing final answers, treating planning, tool use, and code execution as transient interactions. This design is poorly suited to long-horizon bioinformatics tasks, where conclusions must remain connected to the data, computations, and intermediate evidence that support them. We introduce \textbf{Bioinfoysis}, a multi-agent harness that represents each request as a persistent, artifact-grounded analysis run. Bioinfoysis combines global planning with step-wise, evidence-driven replanning: the planner maintains an executable checklist and revises pending steps using structured handoffs returned after each worker execution. These handoffs bind intermediate results to their responsible agent, checklist step, and plan generation, preventing stale evidence from being silently reused after replanning. A controlled runtime validates generated scripts, tables, and figures before they are used in downstream analysis or reporting, while role-specific context, persistent memory, and governed bioinformatics skills support reliable execution over long analysis trajectories. We evaluate Bioinfoysis on BixBench and two question-answering tracks of LAB-Bench 2. On BixBench, Bioinfoysis achieves state-of-the-art accuracy of 82.4\%. Across four underlying language models, Bioinfoysis increases average accuracy from 27.81\% to 64.13\% on SeqQA2 and from 3.13\% to 31.25\% on DbQA2. These results demonstrate that reliable bioinformatics automation depends not only on model capability, but also on the harness that governs planning, execution, memory, and evidence flow. We hope that the emergence of Bioinfoysis will play a driving and leading role in the development of the bioinformatics community. Our demo website can be seen in https://report.bioinfoysis.com/.

cs.AI↗

arXiv2026

An AI-driven robotic system for two-dimensional hetero-assemblies

Nanomaterials stacked on-demand, such as rotationally assembled two-dimensional (2D) van der Waals (vdW) layered compounds, provides a versatile platform for quantum simulation and the exploration of exotic electronic phases. Currently, however, such nanoassemblies remain largely confined to inefficiency, manually operated process, limiting their potential for probing emergent physical phenomena. There is a pressing need in the field for high-precision, automated assembling techniques, especially for the scalable fabrication of 2D twistronic heterostructures. Here, we present an intelligent automation system dedicated to the fabrication of van der Waals stacks, following the state-of-the-art protocol for dry transfer of exfoliated 2D materials. The system further employs metadata generated from each automated stacking procedure to perform reinforcement learning, thereby continuously bettering its performances. As a concrete demonstration, we fabricate twisted bilayer graphene (TBLG) -- known for its challenging preparation -- and exhibit its unconventional superconductivity near the magic angle. Our work may pave the way for high-throughput fabrication of low-dimensional nanomaterials including twistronic heterostructures, where integrating data mining and artificial intelligence can accelerate the discovery of novel physical phenomena.

cond-mat.mes-hall↗

arXiv2025

Scaling Spatial Reasoning in MLLMs through Programmatic Data Synthesis

Embodied intelligence, a grand challenge in artificial intelligence, is fundamentally constrained by the limited spatial understanding and reasoning capabilities of current models. Prevailing efforts to address this through enhancing Vision-Language Models (VLMs) are trapped in a dilemma: template-based datasets are scalable but structurally rigid, while manual annotation is linguistically diverse but unscalable and, critically, computationally imprecise. We introduce SPRITE, a novel framework that overcomes this dilemma by leveraging simulators and large models to programmatically synthesize scalable, diverse, and high-quality spatial reasoning data. The core innovation of SPRITE is to reframe ground-truth generation as a code-generation task. We utilize LLMs to compile complex spatial questions into executable programs, which are then verified against high-precision scene meta-information extracted from simulators. This ensures our ground truth is both computationally precise and verifiable, while the generative power of LLMs provides vast linguistic diversity. Leveraging this pipeline, we have curated a dataset encompassing 3 simulators, 11k+ scenes, and 300k+ image/video instruction-tuning pairs. We demonstrate that a VLM trained on our data achieves significant performance gains on multiple spatial benchmarks and outperforms other open-source datasets of equivalent size. Furthermore, a scalability analysis confirms our hypothesis that overcoming the low-diversity nature of traditional template methods is essential for building robust, generalizable spatial intelligence. We will make the SPRITE framework code and the full 300k+ dataset publicly available to facilitate future research in spatial intelligence.

cs.AI↗

arXiv2025

PVU: Design and Implementation of a Posit Vector Arithmetic Unit (PVU) for Enhanced Floating-Point Computing in Edge and AI Applications

With the rapid development of edge computing, artificial intelligence and other fields, the accuracy and efficiency of floating-point computing have become increasingly crucial. However, the traditional IEEE 754 floating-point system faces bottlenecks in energy consumption and computing accuracy, which have become major constraints. To address this issue, the Posit digital system characterized by adaptive accuracy, broader dynamic range and low hardware consumption has been put forward. Despite its widespread adoption, the existing research mainly concentrates on scalar computation, which is insufficient to meet the requirements of large-scale parallel data processing. This paper proposes, for the first time, a Posit Vector Arithmetic Unit (PVU) designed using the Chisel language. It supports vector operations such as addition, subtraction, multiplication, division, and dot product, thereby overcoming the limitations of traditional scalar designs and integrating the RISC-V instruction extension. The contributions of this paper include the efficient implementation of the vector arithmetic unit, the parametric and modular hardware design as well as the verification of the practical application of the positive digital system. This paper extracts the quantized data of the first convolutional layer for verification. Experiments indicate that the accuracy rate of the division operation is 95.84\%, and the accuracy rate of the remaining operations is 100\%. Moreover, the PVU is implemented with only 65,407 LUTs. Therefore, PVU has great potential as a new-generation floating-point computing platform in various fields.

cs.DC↗

arXiv2023

Uncertainty-aware predictions of molecular X-ray absorption spectra using neural network ensembles

As machine learning (ML) methods continue to be applied to a broad scope of problems in the physical sciences, uncertainty quantification is becoming correspondingly more important for their robust application. Uncertainty aware machine learning methods have been used in select applications, but largely for scalar properties. In this work, we showcase an exemplary study in which neural network ensembles are used to predict the X-ray absorption spectra of small molecules, as well as their point-wise uncertainty, from local atomic environments. The performance of the resulting surrogate clearly demonstrates quantitative correlation between errors relative to ground truth and the predicted uncertainty estimates. Significantly, the model provides an upper bound on the expected error. Specifically, an important quality of this uncertainty-aware model is that it can indicate when the model is predicting on out-of-sample data. This allows for its integration with large scale sampling of structures together with active learning or other techniques for structure refinement. Additionally, our models can be generalized to larger molecules than those used for training, and also successfully track uncertainty due to random distortions in test molecules. While we demonstrate this workflow on a specific example, ensemble learning is completely general. We believe it could have significant impact on ML-enabled forward modeling of a broad array of molecular and materials properties.

cond-mat.mtrl-sci↗

arXiv2023

Decoding Structure-Spectrum Relationships with Physically Organized Latent Spaces

A new semi-supervised machine learning method for the discovery of structure-spectrum relationships is developed and demonstrated using the specific example of interpreting X-ray absorption near-edge structure (XANES) spectra. This method constructs a one-to-one mapping between individual structure descriptors and spectral trends. Specifically, an adversarial autoencoder is augmented with a novel rank constraint (RankAAE). The RankAAE methodology produces a continuous and interpretable latent space, where each dimension can track an individual structure descriptor. As a part of this process, the model provides a robust and quantitative measure of the structure-spectrum relationship by decoupling intertwined spectral contributions from multiple structural characteristics. This makes it ideal for spectral interpretation and the discovery of new descriptors. The capability of this procedure is showcased by considering five local structure descriptors and a database of over fifty thousand simulated XANES spectra across eight first-row transition metal oxide families. The resulting structure-spectrum relationships not only reproduce known trends in the literature, but also reveal unintuitive ones that are visually indiscernible in large data sets. The results suggest that the RankAAE methodology has great potential to assist researchers to interpret complex scientific data, test physical hypotheses, and reveal new patterns that extend scientific insight.

cond-mat.mtrl-sci↗