arXiv subjects

Lili Liu

Publications and source records attributed to Lili Liu.

11 recordsLinked to original sources

arXiv2026

Operator Fusion for LLM Inference on the Tensix Architecture

This study addresses on-device inference bottlenecks of Transformer models on Tenstorrent's Tensix architecture and proposes an operator fusion strategy that enhances data locality. RMSNorm is fused with matrix multiplication in self-attention and in the FFN, enabling back-to-back execution of memory-bound and compute-bound operators in on-chip SRAM to significantly reduce DRAM reads/writes of intermediate results and scheduling overhead. To support multi-core parallelism, a NoC-based multicast mechanism is leveraged in which row/column master nodes efficiently distribute inputs and weights across the core mesh, alleviating DRAM bandwidth contention. Experiments on the Wormhole platform with Qwen2.5-0.5B, Qwen3-0.6B, and Qwen3-4B show up to 37.44% latency reduction for attention and 15.89% for MLP, with up to 7.91% reduction per decoder layer, while Pearson Correlation Coefficient (PCC) remains above 98.75%, confirming significant end-to-end efficiency gains under numerical consistency.

cs.LG↗

arXiv2025

Spatial tangible user interfaces for cognitive assessment and training

This paper discusses Tangible User Interfaces (TUIs) and their potential impact on cognitive assessment and cognitive training. We believe that TUIs, and particularly a subset that we dub spatial TUIs, can extend human computer interaction beyond some of its current limitations. Spatial TUIs exploit human innate spatial and tactile ability in an intuitive and direct manner, affording interaction paradigms that are practically impossible using current interface technology. As proof-of-concept we examine implementations in the field of cognitive assessment and training. In this paper we use Cognitive Cubes, a novel TUI we developed, as an applied test bed for our beliefs, presenting promising experimental results for cognitive assessment of spatial ability, and possibly for training purposes.

cs.HC↗

arXiv2025

Can We Reliably Predict the Fed's Next Move? A Multi-Modal Approach to U.S. Monetary Policy Forecasting

Forecasting central bank policy decisions remains a persistent challenge for investors, financial institutions, and policymakers due to the wide-reaching impact of monetary actions. In particular, anticipating shifts in the U.S. federal funds rate is vital for risk management and trading strategies. Traditional methods relying only on structured macroeconomic indicators often fall short in capturing the forward-looking cues embedded in central bank communications. This study examines whether predictive accuracy can be enhanced by integrating structured data with unstructured textual signals from Federal Reserve communications. We adopt a multi-modal framework, comparing traditional machine learning models, transformer-based language models, and deep learning architectures in both unimodal and hybrid settings. Our results show that hybrid models consistently outperform unimodal baselines. The best performance is achieved by combining TF-IDF features of FOMC texts with economic indicators in an XGBoost classifier, reaching a test AUC of 0.83. FinBERT-based sentiment features marginally improve ranking but perform worse in classification, especially under class imbalance. SHAP analysis reveals that sparse, interpretable features align more closely with policy-relevant signals. These findings underscore the importance of integrating textual and structured signals transparently. For monetary policy forecasting, simpler hybrid models can offer both accuracy and interpretability, delivering actionable insights for researchers and decision-makers.

q-fin.PM↗

arXiv2025

A tangible user interface for assessing cognitive mapping ability

Wayfinding, the ability to recall the environment and navigate through it, is an essential cognitive skill relied upon almost every day in a person's life. A crucial component of wayfinding is the construction of cognitive maps, mental representations of the environments through which a person travels. Age, disease or injury can severely affect cognitive mapping, making assessment of this basic survival skill particularly important to clinicians and therapists. Cognitive mapping has also been the focus of decades of basic research by cognitive psychologists. Both communities have evolved a number of techniques for assessing cognitive mapping ability. We present the Cognitive Map Probe (CMP), a new computerized tool for assessment of cognitive mapping ability that increases consistency and promises improvements in flexibility, accessibility, sensitivity and control. The CMP uses a tangible user interface that affords spatial manipulation. We describe the design of the CMP, and find that it is sensitive to factors known to affect cognitive mapping performance in extensive experimental testing.

cs.HC↗

arXiv2024

Facial Kinship Verification from remote photoplethysmography

Facial Kinship Verification (FKV) aims at automatically determining whether two subjects have a kinship relation based on human faces. It has potential applications in finding missing children and social media analysis. Traditional FKV faces challenges as it is vulnerable to spoof attacks and raises privacy issues. In this paper, we explore for the first time the FKV with vital bio-signals, focusing on remote Photoplethysmography (rPPG). rPPG signals are extracted from facial videos, resulting in a one-dimensional signal that measures the changes in visible light reflection emitted to and detected from the skin caused by the heartbeat. Specifically, in this paper, we employed a straightforward one-dimensional Convolutional Neural Network (1DCNN) with a 1DCNN-Attention module and kinship contrastive loss to learn the kin similarity from rPPGs. The network takes multiple rPPG signals extracted from various facial Regions of Interest (ROIs) as inputs. Additionally, the 1DCNN attention module is designed to learn and capture the discriminative kin features from feature embeddings. Finally, we demonstrate the feasibility of rPPG to detect kinship with the experiment evaluation on the UvANEMO Smile Database from different kin relations.

cs.CV↗

arXiv2023

YAYI 2: Multilingual Open-Source Large Language Models

As the latest advancements in natural language processing, large language models (LLMs) have achieved human-level language understanding and generation abilities in many real-world tasks, and even have been regarded as a potential path to the artificial general intelligence. To better facilitate research on LLMs, many open-source LLMs, such as Llama 2 and Falcon, have recently been proposed and gained comparable performances to proprietary models. However, these models are primarily designed for English scenarios and exhibit poor performances in Chinese contexts. In this technical report, we propose YAYI 2, including both base and chat models, with 30 billion parameters. YAYI 2 is pre-trained from scratch on a multilingual corpus which contains 2.65 trillion tokens filtered by our pre-training data processing pipeline. The base model is aligned with human values through supervised fine-tuning with millions of instructions and reinforcement learning from human feedback. Extensive experiments on multiple benchmarks, such as MMLU and CMMLU, consistently demonstrate that the proposed YAYI 2 outperforms other similar sized open-source models.

cs.CL↗

arXiv2023

Multi-view learning for automatic classification of multi-wavelength auroral images

Auroral classification plays a crucial role in polar research. However, current auroral classification studies are predominantly based on images taken at a single wavelength, typically 557.7 nm. Images obtained at other wavelengths have been comparatively overlooked, and the integration of information from multiple wavelengths remains an underexplored area. This limitation results in low classification rates for complex auroral patterns. Furthermore, these studies, whether employing traditional machine learning or deep learning approaches, have not achieved a satisfactory trade-off between accuracy and speed. To address these challenges, this paper proposes a lightweight auroral multi-wavelength fusion classification network, MLCNet, based on a multi-view approach. Firstly, we develop a lightweight feature extraction backbone, called LCTNet, to improve the classification rate and cope with the increasing amount of auroral observation data. Secondly, considering the existence of multi-scale spatial structures in auroras, we design a novel multi-scale reconstructed feature module named MSRM. Finally, to highlight the discriminative information between auroral classes, we propose a lightweight attention feature enhancement module called LAFE. The proposed method is validated using observational data from the Arctic Yellow River Station during 2003-2004. Experimental results demonstrate that the fusion of multi-wavelength information effectively improves the auroral classification performance. In particular, our approach achieves state-of-the-art classification accuracy compared to previous auroral classification studies, and superior results in terms of accuracy and computational efficiency compared to existing multi-view methods.

cs.CV↗

arXiv2023

Atomic-scale Mapping Unravel Anisotropic Dissolution Behaviors of Gibbsite Nanosheets

This study examines the anisotropic dissolution of the basal plane gibbsite (γ-Al(OH)3) nanoplates in sodium hydroxide solution using in situ atomic force microscopy (AFM) and density functional theory (DFT) calculations. In the surface-reaction controlled regime, in situ AFM measurements reveal anisotropic dissolution of hillocks and etch pits on the gibbsite basal plane, with preferred dissolution directions alternating between layers. The mirror-symmetric pattern of dissolution preference between adjacent gibbsite aluminum hydroxide sheet, observed along the crystallographic a-c plane, results from the matching symmetry between the structures of the adjacent (001) and (002) crystal planes. Consequently, the overall dissolution rate of gibbsite nanoplates exhibits crystallographic a-c plane symmetry, as the rate of parallel steps is governed by the slower ones. DFT calculations suggest that the anisotropic dissolution is partially due to the orientation and strength of Al-OH-Al linkages pair within gibbsite surface structure. These findings offer a comprehensive understanding of anisotropic dissolution behavior of gibbsite and illuminate the mechanisms behind preferential dissolution.

cond-mat.mtrl-sci↗

arXiv2023

Machine learning assisted phase and size-controlled synthesis of iron oxides

The controllable synthesis of iron oxides particles is a critical issue for materials science, energy storage, biomedical applications, environmental science, and earth science. However, synthesis of iron oxides with desired phase and size are still a time-consuming and trial-and-error process. This study presents solutions for two fundamental challenges in materials synthesis: predicting the outcome of a synthesis from specified reaction parameters and correlating sets of parameters to obtain products with desired outcomes. Four machine learning algorithms, including random forest, logistic regression, support vector machine, and k-nearest neighbor, were trained to predict the phase and particle size of iron oxide based on experimental conditions. Among the models, random forest exhibited the best performance, achieving 96% and 81% accuracy when predicting the phase and size of iron oxides in the test dataset. Premutation feature importance analysis shows that most models (except logistic regression) rely on known features such as precursor concentration, pH, and temperature to predict the phases from synthesis conditions. The robustness of the random forest models was further verified by comparing prediction and experimental results based on 24 randomly generated methods in additive and non-additive systems not included in the datasets. The predictions of product phase and particle size from the models are in good agreement with the experimental results. Additionally, a searching and ranking algorithm was developed to recommend potential synthesis parameters for obtaining iron oxide products with desired phase and particle size from previous studies in the dataset.

cond-mat.mtrl-sci↗

arXiv2020

Measurement bias: a structural perspective

The causal structure for measurement bias (MB) remains controversial. Aided by the Directed Acyclic Graph (DAG), this paper proposes a new structure for measuring one singleton variable whose MB arises in the selection of an imperfect I/O device-like measurement system. For effect estimation, however, an extra source of MB arises from any redundant association between a measured exposure and a measured outcome. The misclassification will be bidirectionally differential for a common outcome, unidirectionally differential for a causal relation, and non-differential for a common cause between the measured exposure and the measured outcome or a null effect. The measured exposure can actually affect the measured outcome, or vice versa. Reverse causality is a concept defined at the level of measurement. Our new DAGs have clarified the structures and mechanisms of MB.

stat.ME↗

arXiv2017

A simple and efficient profile likelihood for semiparametric exponential family

Semiparametric exponential family proposed by Ning et al. (2017) is an extension of the parametric exponential family to the case with a nonparametric base measure function. Such a distribution family has potential application in some areas such as high dimensional data analysis. However, the methodology for achieving the semiparametric efficiency has not been proposed in the existing literature. In this paper, we propose a profile likelihood to efficiently estimate both parameter and nonparametric function. Due to the use of the least favorable curve in the procedure of profile likelihood, the semiparametric efficiency is achieved successfully and the estimation bias is reduced significantly. Moreover, by making the most of the structure information of the semiparametric exponential family, the estimator of the least favorable curve has an explicit expression. It ensures that the newly proposed profile likelihood can be implemented and is computationally simple. Simulation studies can illustrate that our proposal is much better than the existing methodology for most cases under study, and is robust to the different model conditions.

stat.ME↗