Source author record

Tam Nguyen

Tam Nguyen appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Machine Learning Artificial Intelligence astro-ph.IM Computation and Language Human-Computer Interaction physics.ins-det Software Engineering Systems and Control

Catalog footprint

What is connected

6works

8topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2026arXiv

Minimizing Collateral Damage in Activation Steering

Activation steering is a method for controlling Large Language Model (LLM) behavior by intervening in its internal representations to increase the alignment with a specific target feature direction. However, standard interventions, such as vector addition, often cause ``collateral damage", defined as unintended changes in the alignment of activations along other non-target feature directions. This damage occurs because standard methods implicitly assume the isotropy of non-target features. In this work, we provide a mathematical formalization of collateral damage and introduce a principled framework that models steering as a constrained optimization problem. Our method finds a new activation that minimizes the expected squared collateral change weighted by the empirical second-moment matrix of activations. This weighting encodes the nonuniform cost of the perturbation in different feature directions, in contrast to isotropic approaches that penalize changes uniformly in all feature directions. By accounting for the empirical second-moment of activations, our approach achieves more precise control while reducing the degradation of model performance on unrelated tasks.

preprint2022arXiv

Improving Transformers with Probabilistic Attention Keys

Multi-head attention is a driving force behind state-of-the-art transformers, which achieve remarkable performance across a variety of natural language processing (NLP) and computer vision tasks. It has been observed that for many applications, those attention heads learn redundant embedding, and most of them can be removed without degrading the performance of the model. Inspired by this observation, we propose Transformer with a Mixture of Gaussian Keys (Transformer-MGK), a novel transformer architecture that replaces redundant heads in transformers with a mixture of keys at each head. These mixtures of keys follow a Gaussian mixture model and allow each attention head to focus on different parts of the input sequence efficiently. Compared to its conventional transformer counterpart, Transformer-MGK accelerates training and inference, has fewer parameters, and requires fewer FLOPs to compute while achieving comparable or better accuracy across tasks. Transformer-MGK can also be easily extended to use with linear attention. We empirically demonstrate the advantage of Transformer-MGK in a range of practical applications, including language modeling and tasks that involve very long sequences. On the Wikitext-103 and Long Range Arena benchmark, Transformer-MGKs with 4 heads attain comparable or better performance to the baseline transformers with 8 heads.

preprint2022arXiv

Transformer with Fourier Integral Attentions

Multi-head attention empowers the recent success of transformers, the state-of-the-art models that have achieved remarkable success in sequence modeling and beyond. These attention mechanisms compute the pairwise dot products between the queries and keys, which results from the use of unnormalized Gaussian kernels with the assumption that the queries follow a mixture of Gaussian distribution. There is no guarantee that this assumption is valid in practice. In response, we first interpret attention in transformers as a nonparametric kernel regression. We then propose the FourierFormer, a new class of transformers in which the dot-product kernels are replaced by the novel generalized Fourier integral kernels. Different from the dot-product kernels, where we need to choose a good covariance matrix to capture the dependency of the features of data, the generalized Fourier integral kernels can automatically capture such dependency and remove the need to tune the covariance matrix. We theoretically prove that our proposed Fourier integral kernels can efficiently approximate any key and query distributions. Compared to the conventional transformers with dot-product attention, FourierFormers attain better accuracy and reduce the redundancy between attention heads. We empirically corroborate the advantages of FourierFormers over the baseline transformers in a variety of practical applications including language modeling and image classification.

preprint2016arXiv

Proof of Control of a UAV and a UGV Cooperating to Manipulate an Object

This paper focuses on the control of a system composed of an Unmanned Aerial Vehicle (UAV) and an Unmanned Ground Vehicle (UGV) which cooperate to manipulate an object. The two units are subject to actuator saturations and cooperate to move the object to a desired pose, characterized by its position and inclination. The paper proposes a control strategy where the ground vehicle is tasked to deploy the object to a certain position, whereas the aerial vehicle adjusts its inclination. The ground vehicle is governed by a saturated proportional-derivative control law. The aerial vehicle is regulated by means of a cascade control specifically designed for this problem that is able to exploit the mechanical interconnection. The stability of the overall system is proved through Input-to-State Stability and Small Gain theorem arguments. To solve the problem of constraints satisfaction, a nonlinear Reference Governor scheme is implemented. Numerical simulations are provided to demonstrate the effectiveness of the proposed method.

preprint2016arXiv

Toward Mining Visual Log of Software

In this paper, we define visual log of a software system as data capturing the interactions between its users and its graphic user interface (GUI), such as screen-shots and screen recordings. We vision that mining such visual log could be useful for bug reproducing and debugging, automated GUI testing, user interface designing, question answering of common usages in software support, etc. Toward that vision, we propose a core framework for mining visual log of software. This framework focuses on detecting GUI elements and changes in visual log, removing users' private data, recognizing user interactions with GUI elements, and learning GUI usage patterns. We also performed a small study on the characteristics of GUI elements in mobile apps. The findings from this study suggested several heuristics to design techniques for recognizing GUI elements and interactions.

preprint2014arXiv

Measurement of the absolute Quantum Efficiency of Hamamatsu model R11410-10 photomultiplier tubes at low temperatures down to liquid xenon boiling point

We report on the measurements of the absolute Quantum Efficiency(QE) for Hamamatsu model R11410-10 PMTs specially designed for the use in low background liquid xenon detectors. QE was measured for five PMTs in a spectral range between 154.5 nm to 400 nm at low temperatures down to -110$^0$C. It was shown that during the PMT cooldown from room temperature to -110 $^0$C (a typical PMT operation temperature in liquid xenon detectors), the absolute QE increases by a factor of 1.1 - 1.15 at 175 nm. The QE growth rate with respect to temperature is wavelength dependent peaking at about 165 nm corresponding to the fastest growth of about -0.07 %QE/$^{0}C$ and at about 200 nm corresponding to slowest growth of below -0.01 %QE/$^{0}C$. A dedicated setup and methods for PMT Quantum Efficiency measurement at low temperatures are described in details.