arXiv subjects

Hua Shen

Publications and source records attributed to Hua Shen.

53 records · Page 3Linked to original sources

arXiv2023

MultiTurnCleanup: A Benchmark for Multi-Turn Spoken Conversational Transcript Cleanup

Current disfluency detection models focus on individual utterances each from a single speaker. However, numerous discontinuity phenomena in spoken conversational transcripts occur across multiple turns, hampering human readability and the performance of downstream NLP tasks. This study addresses these phenomena by proposing an innovative Multi-Turn Cleanup task for spoken conversational transcripts and collecting a new dataset, MultiTurnCleanup1. We design a data labeling schema to collect the high-quality dataset and provide extensive data analysis. Furthermore, we leverage two modeling approaches for experimental evaluation as benchmarks for future research.

cs.CL↗

arXiv2023

Does Human Collaboration Enhance the Accuracy of Identifying LLM-Generated Deepfake Texts?

Advances in Large Language Models (e.g., GPT-4, LLaMA) have improved the generation of coherent sentences resembling human writing on a large scale, resulting in the creation of so-called deepfake texts. However, this progress poses security and privacy concerns, necessitating effective solutions for distinguishing deepfake texts from human-written ones. Although prior works studied humans' ability to detect deepfake texts, none has examined whether "collaboration" among humans improves the detection of deepfake texts. In this study, to address this gap of understanding on deepfake texts, we conducted experiments with two groups: (1) nonexpert individuals from the AMT platform and (2) writing experts from the Upwork platform. The results demonstrate that collaboration among humans can potentially improve the detection of deepfake texts for both groups, increasing detection accuracies by 6.36% for non-experts and 12.76% for experts, respectively, compared to individuals' detection accuracies. We further analyze the explanations that humans used for detecting a piece of text as deepfake text, and find that the strongest indicator of deepfake texts is their lack of coherence and consistency. Our study provides useful insights for future tools and framework designs to facilitate the collaborative human detection of deepfake texts. The experiment datasets and AMT implementations are available at: https://github.com/huashen218/llm-deepfake-human-study.git

cs.CL↗

arXiv2023

Gentopia: A Collaborative Platform for Tool-Augmented LLMs

Augmented Language Models (ALMs) empower large language models with the ability to use tools, transforming them into intelligent agents for real-world interactions. However, most existing frameworks for ALMs, to varying degrees, are deficient in the following critical features: flexible customization, collaborative democratization, and holistic evaluation. We present gentopia, an ALM framework enabling flexible customization of agents through simple configurations, seamlessly integrating various language models, task formats, prompting modules, and plugins into a unified paradigm. Furthermore, we establish gentpool, a public platform enabling the registration and sharing of user-customized agents. Agents registered in gentpool are composable such that they can be assembled together for agent collaboration, advancing the democratization of artificial intelligence. To ensure high-quality agents, gentbench, an integral component of gentpool, is designed to thoroughly evaluate user-customized agents across diverse aspects such as safety, robustness, efficiency, etc. We release gentopia on Github and will continuously move forward.

cs.AI↗

arXiv2023

Parachute: Evaluating Interactive Human-LM Co-writing Systems

A surge of advances in language models (LMs) has led to significant interest in using LMs to build co-writing systems, in which humans and LMs interactively contribute to a shared writing artifact. However, there is a lack of studies assessing co-writing systems in interactive settings. We propose a human-centered evaluation framework, Parachute, for interactive co-writing systems. Parachute showcases an integrative view of interaction evaluation, where each evaluation aspect consists of categorized practical metrics. Furthermore, we present Parachute with a use case to demonstrate how to evaluate and compare co-writing systems using Parachute.

cs.HC↗

arXiv2023

SpeechPrompt v2: Prompt Tuning for Speech Classification Tasks

Prompt tuning is a technology that tunes a small set of parameters to steer a pre-trained language model (LM) to directly generate the output for downstream tasks. Recently, prompt tuning has demonstrated its storage and computation efficiency in both natural language processing (NLP) and speech processing fields. These advantages have also revealed prompt tuning as a candidate approach to serving pre-trained LM for multiple tasks in a unified manner. For speech processing, SpeechPrompt shows its high parameter efficiency and competitive performance on a few speech classification tasks. However, whether SpeechPrompt is capable of serving a large number of tasks is unanswered. In this work, we propose SpeechPrompt v2, a prompt tuning framework capable of performing a wide variety of speech classification tasks, covering multiple languages and prosody-related tasks. The experiment result shows that SpeechPrompt v2 achieves performance on par with prior works with less than 0.15M trainable parameters in a unified framework.

eess.AS↗

arXiv2023

ScatterShot: Interactive In-context Example Curation for Text Transformation

The in-context learning capabilities of LLMs like GPT-3 allow annotators to customize an LLM to their specific tasks with a small number of examples. However, users tend to include only the most obvious patterns when crafting examples, resulting in underspecified in-context functions that fall short on unseen cases. Further, it is hard to know when "enough" examples have been included even for known patterns. In this work, we present ScatterShot, an interactive system for building high-quality demonstration sets for in-context learning. ScatterShot iteratively slices unlabeled data into task-specific patterns, samples informative inputs from underexplored or not-yet-saturated slices in an active learning manner, and helps users label more efficiently with the help of an LLM and the current example set. In simulation studies on two text perturbation scenarios, ScatterShot sampling improves the resulting few-shot functions by 4-5 percentage points over random sampling, with less variance as more examples are added. In a user study, ScatterShot greatly helps users in covering different patterns in the input space and labeling in-context examples more efficiently, resulting in better in-context learning and less user effort.

cs.HC↗

arXiv2022

Too Slow to Be Useful? On Incorporating Humans in the Loop of Smart Speakers

Real-time crowd-powered systems, such as Chorus/Evorus, VizWiz, and Apparition, have shown how incorporating humans into automated systems could supplement where the automatic solutions fall short. However, one unspoken bottleneck of applying such architectures to more scenarios is the longer latency of including humans in the loop of automated systems. For the applications that have hard constraints in turnaround times, human-operated components' longer latency and large speed variation seem to be apparent deal breakers. This paper explicates and quantifies these limitations by using a human-powered text-based backend to hold conversations with users through a voice-only smart speaker. Smart speakers must respond to users' requests within seconds, so the workers behind the scenes only have a few seconds to compose answers. We measured the end-to-end system latency and the conversation quality with eight pairs of participants, showing the challenges and superiority of such systems.

cs.HC↗

arXiv2022

Are Shortest Rationales the Best Explanations for Human Understanding?

Existing self-explaining models typically favor extracting the shortest possible rationales - snippets of an input text "responsible for" corresponding output - to explain the model prediction, with the assumption that shorter rationales are more intuitive to humans. However, this assumption has yet to be validated. Is the shortest rationale indeed the most human-understandable? To answer this question, we design a self-explaining model, LimitedInk, which allows users to extract rationales at any target length. Compared to existing baselines, LimitedInk achieves compatible end-task performance and human-annotated rationale agreement, making it a suitable representation of the recent class of self-explaining models. We use LimitedInk to conduct a user study on the impact of rationale length, where we ask human judges to predict the sentiment label of documents based only on LimitedInk-generated rationales with different lengths. We show rationales that are too short do not help humans predict labels better than randomly masked text, suggesting the need for more careful design of the best human rationales.

cs.CL↗

arXiv2022

Improving fairness in speaker verification via Group-adapted Fusion Network

Modern speaker verification models use deep neural networks to encode utterance audio into discriminative embedding vectors. During the training process, these networks are typically optimized to differentiate arbitrary speakers. This learning process biases the learning of fine voice characteristics towards dominant demographic groups, which can lead to an unfair performance disparity across different groups. This is observed especially with underrepresented demographic groups sharing similar voice characteristics. In this work, we investigate the fairness of speaker verification models on controlled datasets with imbalanced gender distributions, providing direct evidence that model performance suffers for underrepresented groups. To mitigate this disparity we propose the group-adapted fusion network (GFN) architecture, a modular architecture based on group embedding adaptation and score fusion. We show that our method alleviates model unfairness by improving speaker verification both overall and for individual groups. Given imbalanced group representation in training, our proposed method achieves overall equal error rate (EER) reduction of 9.6% to 29.0% relative, reduces minority group EER by 13.7% to 18.6%, and results in 20.0% to 25.4% less EER disparity, compared to baselines. The approach is applicable to other types of training data skew in speaker recognition systems.

eess.AS↗

arXiv2021

A class of ENO schemes with adaptive order for solving hyperbolic conservation laws

We propose a class of essentially non-oscillatory schemes with adaptive order (ENO-AO) for solving hyperbolic conservation laws. The new schemes select candidate stencils by novel smoothness indicators which are the measurements of the minimum discrepancy between the reconstructed polynomials and the neighboring cell averages. The new smoothness indicators measure the smoothness of candidate stencils with unequal sizes in a uniform way, so that we can directly use them to select the optimal stencil from candidates that range from first-order all the way up to the designed high-order. Some benchmark test cases are carried out to demonstrate the accuracy and robustness of the proposed schemes.

math.NA↗

arXiv2021

A rotated characteristic decomposition technique for high-order reconstructions in multi-dimensions

When constructing high-order schemes for solving hyperbolic conservation laws, the corresponding high-order reconstructions are commonly performed in characteristic spaces to eliminate spurious oscillations as much as possible. For multi-dimensional finite volume (FV) schemes, we need to perform the characteristic decomposition several times in different normal directions of the target cell, which is very time-consuming. In this paper, we propose a rotated characteristic decomposition technique which requires only one-time decomposition for multi-dimensional reconstructions. The rotated direction depends only on the gradient of a specific physical quantity which is cheap to calculate. This technique not only reduces the computational cost remarkably, but also controls spurious oscillations effectively. We take a third-order weighted essentially non-oscillatory finite volume (WENO-FV) scheme for solving the Euler equations as an example to demonstrate the efficiency of the proposed technique.

math.NA↗

arXiv2021

A class of high-order weighted compact central schemes for solving hyperbolic conservation laws

We propose a class of weighted compact central (WCC) schemes for solving hyperbolic conservation laws. The linear version can be considered as a high-order extension of the central Lax-Friedrichs (LxF) scheme and the central conservation element and solution element (CESE) scheme. On every cell, the solution is approximated by a Pth order polynomial of which all the DOFs are stored and updated separately. The cell average is updated by a classical finite volume scheme which is constructed based on space-time staggered meshes such that the fluxes are continuous across the interfaces of the adjacent control volumes and, therefore, the local Riemann problem is bypassed. The kth order spatial derivatives are updated by a central difference of (k-1)th order spatial derivatives at cell vertices. All the space-time information is calculated by the Cauchy-Kovalewski procedure. By doing so, the schemes are able to achieve arbitrarily uniform spacetime high order on a super-compact stencil with only one explicit time step. In order to capture discontinuities without spurious oscillations, a weighted essentially non-oscillatory (WENO) type limiter is tailor-made for the schemes. The limiter preserves the compactness and high order accuracy of the schemes. The accuracy, robustness, and efficiency of the schemes are verified by several numerical examples of scalar conservation laws and the compressible Euler equations.

math.NA↗

arXiv2021

Explaining the Road Not Taken

It is unclear if existing interpretations of deep neural network models respond effectively to the needs of users. This paper summarizes the common forms of explanations (such as feature attribution, decision rules, or probes) used in over 200 recent papers about natural language processing (NLP), and compares them against user questions collected in the XAI Question Bank. We found that although users are interested in explanations for the road not taken -- namely, why the model chose one result and not a well-defined, seemly similar legitimate counterpart -- most model interpretations cannot answer these questions.

cs.CL↗

arXiv2021

An alternative reconstruction for WENO schemes with adaptive order

We propose an alternative reconstruction for weighted essentially non-oscillatory schemes with adaptive order (WENO-AO) for solving hyperbolic conservation laws. The alternative reconstruction has a more concise form than the original WENO-AO reconstruction. Moreover, it is a strictly convex combination of polynomials with unequal degrees. Numerical examples show that the alternative reconstruction maintains the accuracy and robustness of the WENO-AO schemes.

math.NA↗

arXiv2020

A Tale of Evil Twins: Adversarial Inputs versus Poisoned Models

Despite their tremendous success in a range of domains, deep learning systems are inherently susceptible to two types of manipulations: adversarial inputs -- maliciously crafted samples that deceive target deep neural network (DNN) models, and poisoned models -- adversely forged DNNs that misbehave on pre-defined inputs. While prior work has intensively studied the two attack vectors in parallel, there is still a lack of understanding about their fundamental connections: what are the dynamic interactions between the two attack vectors? what are the implications of such interactions for optimizing existing attacks? what are the potential countermeasures against the enhanced attacks? Answering these key questions is crucial for assessing and mitigating the holistic vulnerabilities of DNNs deployed in realistic settings. Here we take a solid step towards this goal by conducting the first systematic study of the two attack vectors within a unified framework. Specifically, (i) we develop a new attack model that jointly optimizes adversarial inputs and poisoned models; (ii) with both analytical and empirical evidence, we reveal that there exist intriguing "mutual reinforcement" effects between the two attack vectors -- leveraging one vector significantly amplifies the effectiveness of the other; (iii) we demonstrate that such effects enable a large design spectrum for the adversary to enhance the existing attacks that exploit both vectors (e.g., backdoor attacks), such as maximizing the attack evasiveness with respect to various detection methods; (iv) finally, we discuss potential countermeasures against such optimized attacks and their technical challenges, pointing to several promising research directions.

cs.LG↗

arXiv2020

How Useful Are the Machine-Generated Interpretations to General Users? A Human Evaluation on Guessing the Incorrectly Predicted Labels

Explaining to users why automated systems make certain mistakes is important and challenging. Researchers have proposed ways to automatically produce interpretations for deep neural network models. However, it is unclear how useful these interpretations are in helping users figure out why they are getting an error. If an interpretation effectively explains to users how the underlying deep neural network model works, people who were presented with the interpretation should be better at predicting the model's outputs than those who were not. This paper presents an investigation on whether or not showing machine-generated visual interpretations helps users understand the incorrectly predicted labels produced by image classifiers. We showed the images and the correct labels to 150 online crowd workers and asked them to select the incorrectly predicted labels with or without showing them the machine-generated visual interpretations. The results demonstrated that displaying the visual interpretations did not increase, but rather decreased, the average guessing accuracy by roughly 10%.

cs.HC↗

arXiv2019

Interpretable Deep Learning under Fire

Providing explanations for deep neural network (DNN) models is crucial for their use in security-sensitive domains. A plethora of interpretation models have been proposed to help users understand the inner workings of DNNs: how does a DNN arrive at a specific decision for a given input? The improved interpretability is believed to offer a sense of security by involving human in the decision-making process. Yet, due to its data-driven nature, the interpretability itself is potentially susceptible to malicious manipulations, about which little is known thus far. Here we bridge this gap by conducting the first systematic study on the security of interpretable deep learning systems (IDLSes). We show that existing \imlses are highly vulnerable to adversarial manipulations. Specifically, we present ADV^2, a new class of attacks that generate adversarial inputs not only misleading target DNNs but also deceiving their coupled interpretation models. Through empirical evaluation against four major types of IDLSes on benchmark datasets and in security-critical applications (e.g., skin cancer diagnosis), we demonstrate that with ADV^2 the adversary is able to arbitrarily designate an input's prediction and interpretation. Further, with both analytical and empirical evidence, we identify the prediction-interpretation gap as one root cause of this vulnerability -- a DNN and its interpretation model are often misaligned, resulting in the possibility of exploiting both models simultaneously. Finally, we explore potential countermeasures against ADV^2, including leveraging its low transferability and incorporating it in an adversarial training framework. Our findings shed light on designing and operating IDLSes in a more secure and informative fashion, leading to several promising research directions.

cs.CR↗