Yong Guan

Xiaozhi Wang, Hao Peng, Yong Guan et al. · tsinghua

Understanding events in texts is a core objective of natural language understanding, which requires detecting event occurrences, extracting event arguments, and analyzing inter-event relationships. However, due to the annotation challenges brought by task complexity, a large-scale dataset covering the full process of event understanding has long been absent. In this paper, we introduce MAVEN-Arg, which augments MAVEN datasets with event argument annotations, making the first all-in-one dataset supporting event detection, event argument extraction (EAE), and event relation extraction. As an EAE benchmark, MAVEN-Arg offers three main advantages: (1) a comprehensive schema covering 162 event types and 612 argument roles, all with expert-written definitions and examples; (2) a large data scale, containing 98,591 events and 290,613 arguments obtained with laborious human annotation; (3) the exhaustive annotation supporting all task variants of EAE, which annotates both entity and non-entity event arguments in document level. Experiments indicate that MAVEN-Arg is quite challenging for both fine-tuned EAE models and proprietary large language models (LLMs). Furthermore, to demonstrate the benefits of an all-in-one dataset, we preliminarily explore a potential application, future event prediction, with LLMs. MAVEN-Arg and codes can be obtained from https://github.com/THU-KEG/MAVEN-Argument.

Dong Qin, George Amariucai, Daji Qiao et al.

The limited transparency of the inner decision-making mechanism in deep neural networks (DNN) and other machine learning (ML) models has hindered their application in several domains. In order to tackle this issue, feature attribution methods have been developed to identify the crucial features that heavily influence decisions made by these black box models. However, many feature attribution methods have inherent downsides. For example, one category of feature attribution methods suffers from the artifacts problem, which feeds out-of-distribution masked inputs directly through the classifier that was originally trained on natural data points. Another category of feature attribution method finds explanations by using jointly trained feature selectors and predictors. While avoiding the artifacts problem, this new category suffers from the Encoding Prediction in the Explanation (EPITE) problem, in which the predictor's decisions rely not on the features, but on the masks that selects those features. As a result, the credibility of attribution results is undermined by these downsides. In this research, we introduce the Double-sided Remove and Reconstruct (DoRaR) feature attribution method based on several improvement methods that addresses these issues. By conducting thorough testing on MNIST, CIFAR10 and our own synthetic dataset, we demonstrate that the DoRaR feature attribution method can effectively bypass the above issues and can aid in training a feature selector that outperforms other state-of-the-art feature attribution methods. Our code is available at https://github.com/dxq21/DoRaR.

Zeping Sui, Zilong Liu, Leila Musavian et al.

The impact of both multiplicative and additive hardware impairments (HWIs) on multiple-input multiple-output affine frequency division multiplexing (MIMO-AFDM) systems is investigated. For small-scale MIMO-AFDM systems, a tight bit error rate (BER) upper bound associated with the maximum likelihood (ML) detector is derived. By contrast, for large-scale systems, a closed-form BER approximation associated with the linear minimum mean squared error (LMMSE) detector is presented, including realistic imperfect channel estimation scenarios. Our first key observation is that the full diversity order of a hardware-impaired AFDM system remains unaffected, which is a unique advantage. Furthermore, our analysis shows that 1) the BER results derived accurately predict the simulated ML performance in moderate-to-high signal-to-noise ratios (SNRs), while the theoretical BER curve of the LMMSE detector closely matches that of the Monte-Carlo based one. 2) MIMO-AFDM is more resilient to multiplicative distortions, such as phase noise and carrier frequency offset, compared to its orthogonal frequency division multiplexing (OFDM) counterparts. This is attributed to its inherent chirp signal characteristics; 3) MIMO-AFDM consistently achieves superior BER performance compared to conventional MIMO-OFDM systems under the same additive HWI conditions, as well as different velocity values. The latter is because MIMO-AFDM is also resilient to the additional inter-carrier interference (ICI) imposed by the nonlinear distortions of additive HWIs. In a nutshell, compared to OFDM, AFDM demonstrates stronger ICI resilience and achieves the maximum full diversity attainable gain even under HWIs, thanks to its intrinsic chirp signalling structure as well as to the beneficial spreading effect of the discrete affine Fourier transform.

Yong Guan, Dingxiao Liu, Jinchen Ma et al. · tsinghua

Generative document retrieval, an emerging paradigm in information retrieval, learns to build connections between documents and identifiers within a single model, garnering significant attention. However, there are still two challenges: (1) neglecting inner-content correlation during document representation; (2) lacking explicit semantic structure during identifier construction. Nonetheless, events have enriched relations and well-defined taxonomy, which could facilitate addressing the above two challenges. Inspired by this, we propose Event GDR, an event-centric generative document retrieval model, integrating event knowledge into this task. Specifically, we utilize an exchange-then-reflection method based on multi-agents for event knowledge extraction. For document representation, we employ events and relations to model the document to guarantee the comprehensiveness and inner-content correlation. For identifier construction, we map the events to well-defined event taxonomy to construct the identifiers with explicit semantic structure. Our method achieves significant improvement over the baselines on two datasets, and also hopes to provide insights for future research.