arXiv subjects

Chenguang Wang

Publications and source records attributed to Chenguang Wang.

108 records · Page 6Linked to original sources

arXiv2022

PALT: Parameter-Lite Transfer of Language Models for Knowledge Graph Completion

This paper presents a parameter-lite transfer learning approach of pretrained language models (LM) for knowledge graph (KG) completion. Instead of finetuning, which modifies all LM parameters, we only tune a few new parameters while keeping the original LM parameters fixed. We establish this via reformulating KG completion as a "fill-in-the-blank" task, and introducing a parameter-lite encoder on top of the original LMs. We show that, by tuning far fewer parameters than finetuning, LMs transfer non-trivially to most tasks and reach competitiveness with prior state-of-the-art approaches. For instance, we outperform the fully finetuning approaches on a KG completion benchmark by tuning only 1% of the parameters. The code and datasets are available at \url{https://github.com/yuanyehome/PALT}.

cs.CL↗

arXiv2022

IELM: An Open Information Extraction Benchmark for Pre-Trained Language Models

We introduce a new open information extraction (OIE) benchmark for pre-trained language models (LM). Recent studies have demonstrated that pre-trained LMs, such as BERT and GPT, may store linguistic and relational knowledge. In particular, LMs are able to answer ``fill-in-the-blank'' questions when given a pre-defined relation category. Instead of focusing on pre-defined relations, we create an OIE benchmark aiming to fully examine the open relational information present in the pre-trained LMs. We accomplish this by turning pre-trained LMs into zero-shot OIE systems. Surprisingly, pre-trained LMs are able to obtain competitive performance on both standard OIE datasets (CaRB and Re-OIE2016) and two new large-scale factual OIE datasets (TAC KBP-OIE and Wikidata-OIE) that we establish via distant supervision. For instance, the zero-shot pre-trained LMs outperform the F1 score of the state-of-the-art supervised OIE methods on our factual OIE datasets without needing to use any training sets. Our code and datasets are available at https://github.com/cgraywang/IELM

cs.CL↗

arXiv2022

Fine-mixing: Mitigating Backdoors in Fine-tuned Language Models

Deep Neural Networks (DNNs) are known to be vulnerable to backdoor attacks. In Natural Language Processing (NLP), DNNs are often backdoored during the fine-tuning process of a large-scale Pre-trained Language Model (PLM) with poisoned samples. Although the clean weights of PLMs are readily available, existing methods have ignored this information in defending NLP models against backdoor attacks. In this work, we take the first step to exploit the pre-trained (unfine-tuned) weights to mitigate backdoors in fine-tuned language models. Specifically, we leverage the clean pre-trained weights via two complementary techniques: (1) a two-step Fine-mixing technique, which first mixes the backdoored weights (fine-tuned on poisoned data) with the pre-trained weights, then fine-tunes the mixed weights on a small subset of clean data; (2) an Embedding Purification (E-PUR) technique, which mitigates potential backdoors existing in the word embeddings. We compare Fine-mixing with typical backdoor mitigation methods on three single-sentence sentiment classification tasks and two sentence-pair classification tasks and show that it outperforms the baselines by a considerable margin in all scenarios. We also show that our E-PUR method can benefit existing mitigation methods. Our work establishes a simple but strong baseline defense for secure fine-tuned NLP models against backdoor attacks.

cs.CL↗

arXiv2022

Joint Language Semantic and Structure Embedding for Knowledge Graph Completion

The task of completing knowledge triplets has broad downstream applications. Both structural and semantic information plays an important role in knowledge graph completion. Unlike previous approaches that rely on either the structures or semantics of the knowledge graphs, we propose to jointly embed the semantics in the natural language description of the knowledge triplets with their structure information. Our method embeds knowledge graphs for the completion task via fine-tuning pre-trained language models with respect to a probabilistic structured loss, where the forward pass of the language models captures semantics and the loss reconstructs structures. Our extensive experiments on a variety of knowledge graph benchmarks have demonstrated the state-of-the-art performance of our method. We also show that our method can significantly improve the performance in a low-resource regime, thanks to the better use of semantics. The code and datasets are available at https://github.com/pkusjh/LASS.

cs.CL↗

arXiv2022

A Game-Theoretic Approach for Improving Generalization Ability of TSP Solvers

In this paper, we introduce a two-player zero-sum framework between a trainable \emph{Solver} and a \emph{Data Generator} to improve the generalization ability of deep learning-based solvers for Traveling Salesman Problem (TSP). Grounded in \textsl{Policy Space Response Oracle} (PSRO) methods, our two-player framework outputs a population of best-responding Solvers, over which we can mix and output a combined model that achieves the least exploitability against the Generator, and thereby the most generalizable performance on different TSP tasks. We conduct experiments on a variety of TSP instances with different types and sizes. Results suggest that our Solvers achieve the state-of-the-art performance even on tasks the Solver never meets, whilst the performance of other deep learning-based Solvers drops sharply due to over-fitting. To demonstrate the principle of our framework, we study the learning outcome of the proposed two-player game and demonstrate that the exploitability of the Solver population decreases during training, and it eventually approximates the Nash equilibrium along with the Generator.

cs.LG↗

arXiv2022

Writable spin wave nanochannels in an artificial-spin-ice-mediated ferromagnetic thin film

Magnonics, which employs spin-waves to transmit and process information, is a promising venue for low-power data processing. One of the major challenges is the local control of the spin-wave propagation path. Here, we introduce the concept of writable magnonics by taking advantage of the highly flexible reconfigurability and rewritability of artificial spin ice systems. Using micromagnetic simulations, we show that globally switchable spin-wave propagation and the locally writable spin-wave nanochannels can be realized in a ferromagnetic thin film underlying an artificial pinwheel spin ice. The rewritable magnonics enabled by reconfigurable spin wave nanochannels provides a unique setting to design programmable magnonic circuits and logic devices for ultra-low power applications.

cond-mat.mes-hall↗

arXiv2021

Protecting Intellectual Property of Language Generation APIs with Lexical Watermark

Nowadays, due to the breakthrough in natural language generation (NLG), including machine translation, document summarization, image captioning, etc NLG models have been encapsulated in cloud APIs to serve over half a billion people worldwide and process over one hundred billion word generations per day. Thus, NLG APIs have already become essential profitable services in many commercial companies. Due to the substantial financial and intellectual investments, service providers adopt a pay-as-you-use policy to promote sustainable market growth. However, recent works have shown that cloud platforms suffer from financial losses imposed by model extraction attacks, which aim to imitate the functionality and utility of the victim services, thus violating the intellectual property (IP) of cloud APIs. This work targets at protecting IP of NLG APIs by identifying the attackers who have utilized watermarked responses from the victim NLG APIs. However, most existing watermarking techniques are not directly amenable for IP protection of NLG APIs. To bridge this gap, we first present a novel watermarking method for text generation APIs by conducting lexical modification to the original outputs. Compared with the competitive baselines, our watermark approach achieves better identifiable performance in terms of p-value, with fewer semantic losses. In addition, our watermarks are more understandable and intuitive to humans than the baselines. Finally, the empirical studies show our approach is also applicable to queries from different domains, and is effective on the attacker trained on a mixture of the corpus which includes less than 10\% watermarked samples.

cs.CR↗

arXiv2021

Zero-Shot Information Extraction as a Unified Text-to-Triple Translation

We cast a suite of information extraction tasks into a text-to-triple translation framework. Instead of solving each task relying on task-specific datasets and models, we formalize the task as a translation between task-specific input text and output triples. By taking the task-specific input, we enable a task-agnostic translation by leveraging the latent knowledge that a pre-trained language model has about the task. We further demonstrate that a simple pre-training task of predicting which relational information corresponds to which input text is an effective way to produce task-specific outputs. This enables the zero-shot transfer of our framework to downstream tasks. We study the zero-shot performance of this framework on open information extraction (OIE2016, NYT, WEB, PENN), relation classification (FewRel and TACRED), and factual probe (Google-RE and T-REx). The model transfers non-trivially to most tasks and is often competitive with a fully supervised method without the need for any task-specific training. For instance, we significantly outperform the F1 score of the supervised open information extraction without needing to use its training set.

cs.CL↗

arXiv2021

Learning Graph Representation by Aggregating Subgraphs via Mutual Information Maximization

In this paper, we introduce a self-supervised learning method to enhance the graph-level representations with the help of a set of subgraphs. For this purpose, we propose a universal framework to generate subgraphs in an auto-regressive way and then using these subgraphs to guide the learning of graph representation by Graph Neural Networks. Under this framework, we can get a comprehensive understanding of the graph structure in a learnable way. And to fully capture enough information of original graphs, we design three information aggregators: \textbf{attribute-conv}, \textbf{layer-conv} and \textbf{subgraph-conv} to gather information from different aspects. And to achieve efficient and effective contrastive learning, a Head-Tail contrastive construction is proposed to provide abundant negative samples. Under all proposed components which can be generalized to any Graph Neural Networks, in the unsupervised case, we achieve new state-of-the-art results in several benchmarks. We also evaluate our model on semi-supervised learning tasks and make a fair comparison to state-of-the-art semi-supervised methods.

cs.LG↗

arXiv2021

Tailoring magnetization reversal of a single-domain bar nanomagnet via its end geometry

Nanoscale single-domain bar magnets are building blocks for a variety of fundamental and applied mesoscopic magnetic systems, such as artificial spin ices, magnetic shape-morphing microbots as well as magnetic majority logic gates. The magnetization reversal switching field of the bar nanomagnets is a crucial parameter that determines the physical properties and functionalities of their constituted artificial systems. Previous methods on tuning the magnetization reversal switching field of a bar nanomagnet usually rely on modifying its aspect ratio, such as its length, width and/or thickness. Here, we show that the switching field of a bar nanomagnet saturates when extending its length beyond a certain value, preventing further tailoring of the magnetization reversal via aspect ratios. We showcase highly tunable switching field of a bar nanomagent by tailoring its end geometry without altering its size. This provides an easy method to control the magnetization reversal of a single-domain bar nanomagnet. It would enable new research and/or applications, such as designing artificial spin ices with additional tuning parameters, engineering magnetic microbots with more flexibility as well as developing magnetic quantum-dot cellular automata systems for low power computing.

cond-mat.mes-hall↗

arXiv2020

PoD: Positional Dependency-Based Word Embedding for Aspect Term Extraction

Dependency context-based word embedding jointly learns the representations of word and dependency context, and has been proved effective in aspect term extraction. In this paper, we design the positional dependency-based word embedding (PoD) which considers both dependency context and positional context for aspect term extraction. Specifically, the positional context is modeled via relative position encoding. Besides, we enhance the dependency context by integrating more lexical information (e.g., POS tags) along dependency paths. Experiments on SemEval 2014/2015/2016 datasets show that our approach outperforms other embedding methods in aspect term extraction.

cs.CL↗

arXiv2020

Language Models are Open Knowledge Graphs

This paper shows how to construct knowledge graphs (KGs) from pre-trained language models (e.g., BERT, GPT-2/3), without human supervision. Popular KGs (e.g, Wikidata, NELL) are built in either a supervised or semi-supervised manner, requiring humans to create knowledge. Recent deep language models automatically acquire knowledge from large-scale corpora via pre-training. The stored knowledge has enabled the language models to improve downstream NLP tasks, e.g., answering questions, and writing code and articles. In this paper, we propose an unsupervised method to cast the knowledge contained within language models into KGs. We show that KGs are constructed with a single forward pass of the pre-trained language models (without fine-tuning) over the corpora. We demonstrate the quality of the constructed KGs by comparing to two KGs (Wikidata, TAC KBP) created by humans. Our KGs also provide open factual knowledge that is new in the existing KGs. Our code and KGs will be made publicly available.

cs.CL↗

arXiv2020

Transformer on a Diet

Transformer has been widely used thanks to its ability to capture sequence information in an efficient way. However, recent developments, such as BERT and GPT-2, deliver only heavy architectures with a focus on effectiveness. In this paper, we explore three carefully-designed light Transformer architectures to figure out whether the Transformer with less computations could produce competitive results. Experimental results on language model benchmark datasets hint that such trade-off is promising, and the light Transformer reduces 70% parameters at best, while obtains competitive perplexity compared to standard Transformer. The source code is publicly available.

cs.CL↗

arXiv2020

GluonCV and GluonNLP: Deep Learning in Computer Vision and Natural Language Processing

We present GluonCV and GluonNLP, the deep learning toolkits for computer vision and natural language processing based on Apache MXNet (incubating). These toolkits provide state-of-the-art pre-trained models, training scripts, and training logs, to facilitate rapid prototyping and promote reproducible research. We also provide modular APIs with flexible building blocks to enable efficient customization. Leveraging the MXNet ecosystem, the deep learning models in GluonCV and GluonNLP can be deployed onto a variety of platforms with different programming languages. The Apache 2.0 license has been adopted by GluonCV and GluonNLP to allow for software distribution, modification, and usage.

cs.LG↗

arXiv2019

Meta-Path Constrained Random Walk Inference for Large-Scale Heterogeneous Information Networks

Heterogeneous information network (HIN) has shown its power of modeling real world data as a multi-typed entity-relation graph. Meta-path is the key contributor to this power since it enables inference by capturing the proximities between entities via rich semantic links. Previous HIN studies ask users to provide either 1) the meta-path(s) directly or 2) biased examples to generate the meta-path(s). However, lots of HINs (e.g., YAGO2 and Freebase) have rich schema consisting of a sophisticated and large number of types of entities and relations. It is impractical for users to provide the meta-path(s) to support the large scale inference, and biased examples will result in incorrect meta-path based inference, thus limit the power of the meta-path. In this paper, we propose a meta-path constrained inference framework to further release the ability of the meta-path, by efficiently learning the HIN inference patterns via a carefully designed tree structure; and performing unbiased random walk inference with little user guidance. The experiment results on YAGO2 and DBLP datasets show the state-of-the-art performance of the meta-path constrained inference framework.

cs.SI↗

arXiv2019

Language Models with Transformers

The Transformer architecture is superior to RNN-based models in computational efficiency. Recently, GPT and BERT demonstrate the efficacy of Transformer models on various NLP tasks using pre-trained language models on large-scale corpora. Surprisingly, these Transformer architectures are suboptimal for language model itself. Neither self-attention nor the positional encoding in the Transformer is able to efficiently incorporate the word-level sequential context crucial to language modeling. In this paper, we explore effective Transformer architectures for language model, including adding additional LSTM layers to better capture the sequential context while still keeping the computation efficient. We propose Coordinate Architecture Search (CAS) to find an effective architecture through iterative refinement of the model. Experimental results on the PTB, WikiText-2, and WikiText-103 show that CAS achieves perplexities between 20.42 and 34.11 on all problems, i.e. on average an improvement of 12.0 perplexity units compared to state-of-the-art LSTMs. The source code is publicly available.

cs.CL↗

arXiv2019

Percentile-Based Residuals for Model Assessment

Residuals are a key component of diagnosing model fit. The usual practice is to compute standardized residuals using expected values and standard deviations of the observed data, then use these values to detect outliers and assess model fit. Approximate normality of these residuals is key for this process to have good properties, but in many modeling contexts, especially for complex, multi-level models, normality may not hold. In these cases outlier detection and model diagnostics aren't properly calibrated. Alternatively, as we demonstrate, residuals computed from the percentile location of a datum's value in its full predictive distribution lead to well calibrated evaluations of model fit. We generalize an approach described by Dunn and Smyth (1996) and evaluate properties mathematically, via case-studies and by simulation. In addition, we show that the standard residuals can be calibrated to mimic the percentile approach, but that this extra step is avoided by directly using percentile-based residuals. For both the percentile-based residuals and the calibrated standard residuals, the use of full predictive distributions with the appropriate location, spread and shape is necessary for valid assessments.

stat.ME↗

arXiv2016

World Knowledge as Indirect Supervision for Document Clustering

One of the key obstacles in making learning protocols realistic in applications is the need to supervise them, a costly process that often requires hiring domain experts. We consider the framework to use the world knowledge as indirect supervision. World knowledge is general-purpose knowledge, which is not designed for any specific domain. Then the key challenges are how to adapt the world knowledge to domains and how to represent it for learning. In this paper, we provide an example of using world knowledge for domain dependent document clustering. We provide three ways to specify the world knowledge to domains by resolving the ambiguity of the entities and their types, and represent the data with world knowledge as a heterogeneous information network. Then we propose a clustering algorithm that can cluster multiple types and incorporate the sub-type information as constraints. In the experiments, we use two existing knowledge bases as our sources of world knowledge. One is Freebase, which is collaboratively collected knowledge about entities and their organizations. The other is YAGO2, a knowledge base automatically extracted from Wikipedia and maps knowledge to the linguistic knowledge base, WordNet. Experimental results on two text benchmark datasets (20newsgroups and RCV1) show that incorporating world knowledge as indirect supervision can significantly outperform the state-of-the-art clustering algorithms as well as clustering algorithms enhanced with world knowledge features.

cs.LG↗