Large language models (LLMs) have shown impressive abilities for open-domain NLP tasks. However, LLMs are sometimes too footloose for natural language understanding (NLU) tasks which always have restricted output and input format. Their performances on NLU tasks are highly related to prompts or demonstrations and are shown to be poor at performing several representative NLU tasks, such as event extraction and entity typing. To this end, we present SeqGPT, a bilingual (i.e., English and Chinese) open-source autoregressive model specially enhanced for open-domain natural language understanding. We express all NLU tasks with two atomic tasks, which define fixed instructions to restrict the input and output format but still ``open'' for arbitrarily varied label sets. The model is first instruction-tuned with extremely fine-grained labeled data synthesized by ChatGPT and then further fine-tuned by 233 different atomic tasks from 152 datasets across various domains. The experimental results show that SeqGPT has decent classification and extraction ability, and is capable of performing language understanding tasks on unseen domains. We also conduct empirical studies on the scaling of data and model size as well as on the transfer across tasks. Our models are accessible at https://github.com/Alibaba-NLP/SeqGPT.

Microgrid is a typical cyber-physical system which integrates the power flow and the communication flow. Privacy preservation of sensitive information delivering over the communication network has received significant attention. In this paper, a privacy-preserving distributed secondary controller is proposed for DC microgrids. The privacy information is subjected to an output mask function to ensure its anonymity, thereby providing real-time security for private power information. Furthermore, by employing event-triggered mechanism to determine the updating of control signals, current sharing and voltage regulation are achieved simultaneously. In addition, the decreasing event-triggered threshold improves the systems response speed and accuracy. Finally, the simulation and experimental cases demonstrate the effectiveness of the proposed controller.

Vertical-looking radar (VLR) is a significant milestone in the development of insect radars with the capability of detecting the behavior of migratory insects and their biological parameters. In current VLRs, high-speed continuous sampling and long-time integration can barely be performed simultaneously, leading to a low detection probability for tiny insects (weight < 10 mg). Based on the large amount of data acquired by our developed high-range resolution insect radar, the insect echo signals and vertical motion characteristics are initially analyzed and demonstrate that the linear-motion mode is dominant in insect migration; also, the echo signal power of most insects follows the gamma distribution. Based on these characteristics, a long-time integration and detection method for detecting migratory insects, especially tiny targets from echo signals that often dip below the noise level, is proposed. The radial target velocity is also measured as one of the output parameters. The theoretical derivation and optimal choice of detection thresholds are also presented. Simulation and experimental results demonstrate that the proposed method exhibits better insect detection performance and effectively increases the detection range compared with conventional methods. In addition, the measured target velocity can be directly applied to current continuous-sampling VLRs for the ascent and descent rate analysis. Many typical insect migration phenomena have been detected effectively utilizing our developed VLR, and the measured ascent and descent rates of insects agree well with typical take-off, cruising, and landing behaviors. This is the first reported successful VLR application on take-off and landing behaviors of migratory tiny and dense insects.

Semantic dependency parsing, which aims to find rich bi-lexical relationships, allows words to have multiple dependency heads, resulting in graph-structured representations.We propose an approach to semi-supervised learning of semantic dependency parsers based on the CRF autoencoder framework.Our encoder is a discriminative neural semantic dependency parser that predicts the latent parse graph of the input sentence.Our decoder is a generative neural model that reconstructs the input sentence conditioned on the latent parse graph.Our model is arc-factored and therefore parsing and learning are both tractable.Experiments show our model achieves significant and consistent improvement over the supervised baseline.

Unsupervised constituency parsing aims to learn a constituency parser from a training corpus without parse tree annotations.While many methods have been proposed to tackle the problem, including statistical and neural methods, their experimental results are often not directly comparable due to discrepancies in datasets, data preprocessing, lexicalization, and evaluation metrics.In this paper, we first examine experimental settings used in previous work and propose to standardize the settings for better comparability between methods.We then empirically compare several existing methods, including decade-old and newly proposed ones, under the standardized settings on English and Japanese, two languages with different branching tendencies.We find that recent models do not show a clear advantage over decade-old models in our experiments.We hope our work can provide new insights into existing methods and facilitate future empirical evaluation of unsupervised constituency parsing.

Deep inside-outside recursive autoencoder (DIORA) is a neural-based model designed for unsupervised constituency parsing.During its forward computation, it provides phrase and contextual representations for all spans in the input sentence.By utilizing the contextual representation of each leaf-level span, the span of length 1, to reconstruct the word inside the span, the model is trained without labeled data.In this work, we extend the training objective of DIORA by making use of all spans instead of only leaf-level spans.We test our new training objective on datasets of two languages: English and Japanese, and empirically show that our method achieves improvement in parsing accuracy over the original DIORA.

Unsupervised dependency parsing, which tries to discover linguistic dependency structures from unannotated data, is a very challenging task.Almost all previous work on this task focuses on learning generative models.In this paper, we develop an unsupervised dependency parsing model based on the CRF autoencoder.The encoder part of our model is discriminative and globally normalized which allows us to use rich features as well as universal linguistic priors.We propose an exact algorithm for parsing as well as a tractable learning algorithm.We evaluated the performance of our model on eight multilingual treebanks and found that our model achieved comparable performance with state-of-the-art approaches.