Cong Liu


2024

pdf bib
ChatMusician: Understanding and Generating Music Intrinsically with LLM
Ruibin Yuan | Hanfeng Lin | Yi Wang | Zeyue Tian | Shangda Wu | Tianhao Shen | Ge Zhang | Yuhang Wu | Cong Liu | Ziya Zhou | Liumeng Xue | Ziyang Ma | Qin Liu | Tianyu Zheng | Yizhi Li | Yinghao Ma | Yiming Liang | Xiaowei Chi | Ruibo Liu | Zili Wang | Chenghua Lin | Qifeng Liu | Tao Jiang | Wenhao Huang | Wenhu Chen | Jie Fu | Emmanouil Benetos | Gus Xia | Roger Dannenberg | Wei Xue | Shiyin Kang | Yike Guo
Findings of the Association for Computational Linguistics: ACL 2024

While LLMs demonstrate impressive capabilities in musical knowledge, we find that music reasoning is still an unsolved task.We introduce ChatMusician, an open-source large language model (LLM) that integrates intrinsic musical abilities. It is based on continual pre-training and finetuning LLaMA2 on a text-compatible music representation, ABC notation, and the music is treated as a second language.ChatMusician can understand and generate music with a pure text tokenizer without external multi-modal neural structures or tokenizers. Interestingly, endowing musical abilities does not harm language abilities, even achieving a slightly higher MMLU score.ChatMusician is capable of composing well-structured, full-length music, condition on texts, chords, melodies, motifs, musical forms, etc.On our meticulously curated college-level music understanding benchmark, MusicTheoryBench, ChatMusician surpasses LLaMA2 and GPT-3.5 by a noticeable margin. We show that ChatMusician preserves or even surpasses the original LLaMA2 7B’s language abilities by evaluating on MMLU benchmark.Our work reveals that LLMs can be an excellent compressor for music, which can be seen as humanity’s creative language, but there remains significant territory to be conquered.We release our 5B token music-language corpora MusicPiles, the collected MusicTheoryBench, code, model and demo.

pdf bib
Distantly-Supervised Joint Extraction with Noise-Robust Learning
Yufei Li | Xiao Yu | Yanghong Guo | Yanchi Liu | Haifeng Chen | Cong Liu
Findings of the Association for Computational Linguistics: ACL 2024

Joint entity and relation extraction is a process that identifies entity pairs and their relations using a single model. We focus on the problem of joint extraction in distantly-labeled data, whose labels are generated by aligning entity mentions with the corresponding entity and relation tags using a knowledge base (KB). One key challenge is the presence of noisy labels arising from both incorrect entity and relation annotations, which significantly impairs the quality of supervised learning. Existing approaches, either considering only one source of noise or making decisions using external knowledge, cannot well-utilize significant information in the training data. We propose DENRL, a generalizable framework that 1) incorporates a lightweight transformer backbone into a sequence labeling scheme for joint tagging, and 2) employs a noise-robust framework that regularizes the tagging model with significant relation patterns and entity-relation dependencies, then iteratively self-adapts to instances with less noise from both sources. Surprisingly, experiments on two benchmark datasets show that DENRL, using merely its own parametric distribution and simple data-driven heuristics, outperforms strong baselines by a large margin with better interpretability.

pdf bib
Safety Alignment in NLP Tasks: Weakly Aligned Summarization as an In-Context Attack
Yu Fu | Yufei Li | Wen Xiao | Cong Liu | Yue Dong
Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Recent developments in balancing the usefulness and safety of Large Language Models (LLMs) have raised a critical question: Are mainstream NLP tasks adequately aligned with safety consideration? Our study, focusing on safety-sensitive documents obtained through adversarial attacks, reveals significant disparities in the safety alignment of various NLP tasks. For instance, LLMs can effectively summarize malicious long documents but often refuse to translate them. This discrepancy highlights a previously unidentified vulnerability: attacks exploiting tasks with weaker safety alignment, like summarization, can potentially compromise the integrity of tasks traditionally deemed more robust, such as translation and question-answering (QA). Moreover, the concurrent use of multiple NLP tasks with lesser safety alignment increases the risk of LLMs inadvertently processing harmful content. We demonstrate these vulnerabilities in various safety-aligned LLMs, particularly Llama2 models, Gemini and GPT-4, indicating an urgent need for strengthening safety alignments across a broad spectrum of NLP tasks.

2023

pdf bib
White-Box Multi-Objective Adversarial Attack on Dialogue Generation
Yufei Li | Zexin Li | Yingfan Gao | Cong Liu
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Pre-trained transformers are popular in state-of-the-art dialogue generation (DG) systems. Such language models are, however, vulnerable to various adversarial samples as studied in traditional tasks such as text classification, which inspires our curiosity about their robustness in DG systems. One main challenge of attacking DG models is that perturbations on the current sentence can hardly degrade the response accuracy because the unchanged chat histories are also considered for decision-making. Instead of merely pursuing pitfalls of performance metrics such as BLEU, ROUGE, we observe that crafting adversarial samples to force longer generation outputs benefits attack effectiveness—the generated responses are typically irrelevant, lengthy, and repetitive. To this end, we propose a white-box multi-objective attack method called DGSlow. Specifically, DGSlow balances two objectives—generation accuracy and length, via a gradient-based multi-objective optimizer and applies an adaptive searching mechanism to iteratively craft adversarial samples with only a few modifications. Comprehensive experiments on four benchmark datasets demonstrate that DGSlow could significantly degrade state-of-the-art DG models with a higher success rate than traditional accuracy-based methods. Besides, our crafted sentences also exhibit strong transferability in attacking other models.

pdf bib
Dynamic Transformers Provide a False Sense of Efficiency
Yiming Chen | Simin Chen | Zexin Li | Wei Yang | Cong Liu | Robby Tan | Haizhou Li
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Despite much success in natural language processing (NLP), pre-trained language models typically lead to a high computational cost during inference. Multi-exit is a mainstream approach to address this issue by making a trade-off between efficiency and accuracy, where the saving of computation comes from an early exit. However, whether such saving from early-exiting is robust remains unknown. Motivated by this, we first show that directly adapting existing adversarial attack approaches targeting model accuracy cannot significantly reduce inference efficiency. To this end, we propose a simple yet effective attacking framework, SAME, a novel slowdown attack framework on multi-exit models, which is specially tailored to reduce the efficiency of the multi-exit models. By leveraging the multi-exit models’ design characteristics, we utilize all internal predictions to guide the adversarial sample generation instead of merely considering the final prediction. Experiments on the GLUE benchmark show that SAME can effectively diminish the efficiency gain of various multi-exit models by 80% on average, convincingly validating its effectiveness and generalization ability.

pdf bib
GIFT: Graph-Induced Fine-Tuning for Multi-Party Conversation Understanding
Jia-Chen Gu | Zhenhua Ling | Quan Liu | Cong Liu | Guoping Hu
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Addressing the issues of who saying what to whom in multi-party conversations (MPCs) has recently attracted a lot of research attention. However, existing methods on MPC understanding typically embed interlocutors and utterances into sequential information flows, or utilize only the superficial of inherent graph structures in MPCs. To this end, we present a plug-and-play and lightweight method named graph-induced fine-tuning (GIFT) which can adapt various Transformer-based pre-trained language models (PLMs) for universal MPC understanding. In detail, the full and equivalent connections among utterances in regular Transformer ignore the sparse but distinctive dependency of an utterance on another in MPCs. To distinguish different relationships between utterances, four types of edges are designed to integrate graph-induced signals into attention mechanisms to refine PLMs originally designed for processing sequential texts. We evaluate GIFT by implementing it into three PLMs, and test the performance on three downstream tasks including addressee recognition, speaker identification and response selection. Experimental results show that GIFT can significantly improve the performance of three PLMs on three downstream tasks and two benchmarks with only 4 additional parameters per encoding layer, achieving new state-of-the-art performance on MPC understanding.

pdf bib
Uncertainty-Aware Bootstrap Learning for Joint Extraction on Distantly-Supervised Data
Yufei Li | Xiao Yu | Yanchi Liu | Haifeng Chen | Cong Liu
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)

Jointly extracting entity pairs and their relations is challenging when working on distantly-supervised data with ambiguous or noisy labels. To mitigate such impact, we propose uncertainty-aware bootstrap learning, which is motivated by the intuition that the higher uncertainty of an instance, the more likely the model confidence is inconsistent with the ground truths. Specifically, we first explore instance-level data uncertainty to create an initial high-confident examples. Such subset serves as filtering noisy instances and facilitating the model to converge fast at the early stage. During bootstrap learning, we propose self-ensembling as a regularizer to alleviate inter-model uncertainty produced by noisy labels. We further define probability variance of joint tagging probabilities to estimate inner-model parametric uncertainty, which is used to select and build up new reliable training instances for the next iteration. Experimental results on two large datasets reveal that our approach outperforms existing strong baselines and related methods.

pdf bib
MADNet: Maximizing Addressee Deduction Expectation for Multi-Party Conversation Generation
Jia-Chen Gu | Chao-Hong Tan | Caiyuan Chu | Zhen-Hua Ling | Chongyang Tao | Quan Liu | Cong Liu
Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing

Modeling multi-party conversations (MPCs) with graph neural networks has been proven effective at capturing complicated and graphical information flows. However, existing methods rely heavily on the necessary addressee labels and can only be applied to an ideal setting where each utterance must be tagged with an “@” or other equivalent addressee label. To study the scarcity of addressee labels which is a common issue in MPCs, we propose MADNet that maximizes addressee deduction expectation in heterogeneous graph neural networks for MPC generation. Given an MPC with a few addressee labels missing, existing methods fail to build a consecutively connected conversation graph, but only a few separate conversation fragments instead. To ensure message passing between these conversation fragments, four additional types of latent edges are designed to complete a fully-connected graph. Besides, to optimize the edge-type-dependent message passing for those utterances without addressee labels, an Expectation-Maximization-based method that iteratively generates silver addressee labels (E step), and optimizes the quality of generated responses (M step), is designed. Experimental results on two Ubuntu IRC channel benchmarks show that MADNet outperforms various baseline models on the task of MPC generation, especially under the more common and challenging setting where part of addressee labels are missing.

pdf bib
Exploring Prompt-based Multi-task Learning for Multimodal Dialog State Tracking and Immersive Multimodal Conversation
Yirong Chen | Ya Li | Tao Wang | Xiaofen Xing | Xiangmin Xu | Quan Liu | Cong Liu | Guoping Hu
Proceedings of The Eleventh Dialog System Technology Challenge

With the rise of the metaverse, immersive multimodal conversation has attracted more and more researchers’ attention. Multimodal contexts will become more important for human-computer interaction in the metaverse, especially in shopping domain. Unlike traditional conversation tasks, immersive multimodal conversation has challenges such as multimodal ambiguous candidate identification and multimodal coreference resolution, which makes it more difficult to dialog state tracking and response generation, as described in SIMMC 2.1 challenge, a part of DSTC11. In particular, as the number of objects in the scene increases, the difficulty will increase dramatically. We proposed a prompt-based multi-task learning Encoder-Decoder, in which different subtasks use different prompts to make the model tend to focus on the current subtask. We achieve the winner in ambiguous candidates indentification and runner-up in multimodal coreference resolution (MM-Coref), multimodal dialog state tracking (MM-DST) and assistant response generation. Our code and model are made publicly available at https://github.com/scutcyr/dstc11-simmc2.1-scut-bds-lab.

2022

pdf bib
Wider & Closer: Mixture of Short-channel Distillers for Zero-shot Cross-lingual Named Entity Recognition
Jun-Yu Ma | Beiduo Chen | Jia-Chen Gu | Zhenhua Ling | Wu Guo | Quan Liu | Zhigang Chen | Cong Liu
Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing

Zero-shot cross-lingual named entity recognition (NER) aims at transferring knowledge from annotated and rich-resource data in source languages to unlabeled and lean-resource data in target languages. Existing mainstream methods based on the teacher-student distillation framework ignore the rich and complementary information lying in the intermediate layers of pre-trained language models, and domain-invariant information is easily lost during transfer. In this study, a mixture of short-channel distillers (MSD) method is proposed to fully interact the rich hierarchical information in the teacher model and to transfer knowledge to the student model sufficiently and efficiently. Concretely, a multi-channel distillation framework is designed for sufficient information transfer by aggregating multiple distillers as a mixture. Besides, an unsupervised method adopting parallel domain adaptation is proposed to shorten the channels between the teacher and student models to preserve domain-invariant features. Experiments on four datasets across nine languages demonstrate that the proposed method achieves new state-of-the-art performance on zero-shot cross-lingual NER and shows great generalization and compatibility across languages and fields.

2016

pdf bib
Semantic Relation Classification via Hierarchical Recurrent Neural Network with Attention
Minguang Xiao | Cong Liu
Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers

Semantic relation classification remains a challenge in natural language processing. In this paper, we introduce a hierarchical recurrent neural network that is capable of extracting information from raw sentences for relation classification. Our model has several distinctive features: (1) Each sentence is divided into three context subsequences according to two annotated nominals, which allows the model to encode each context subsequence independently so as to selectively focus as on the important context information; (2) The hierarchical model consists of two recurrent neural networks (RNNs): the first one learns context representations of the three context subsequences respectively, and the second one computes semantic composition of these three representations and produces a sentence representation for the relationship classification of the two nominals. (3) The attention mechanism is adopted in both RNNs to encourage the model to concentrate on the important information when learning the sentence representations. Experimental results on the SemEval-2010 Task 8 dataset demonstrate that our model is comparable to the state-of-the-art without using any hand-crafted features.

pdf bib
Improved Word Embeddings with Implicit Structure Information
Jie Shen | Cong Liu
Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers

Distributed word representation is an efficient method for capturing semantic and syntactic word relations. In this work, we introduce an extension to the continuous bag-of-words model for learning word representations efficiently by using implicit structure information. Instead of relying on a syntactic parser which might be noisy and slow to build, we compute weights representing probabilities of syntactic relations based on the Huffman softmax tree in an efficient heuristic. The constructed “implicit graphs” from these weights show that these weights contain useful implicit structure information. Extensive experiments performed on several word similarity and word analogy tasks show gains compared to the basic continuous bag-of-words model.

pdf bib
Modelling Sentence Pairs with Tree-structured Attentive Encoder
Yao Zhou | Cong Liu | Yan Pan
Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers

We describe an attentive encoder that combines tree-structured recursive neural networks and sequential recurrent neural networks for modelling sentence pairs. Since existing attentive models exert attention on the sequential structure, we propose a way to incorporate attention into the tree topology. Specially, given a pair of sentences, our attentive encoder uses the representation of one sentence, which generated via an RNN, to guide the structural encoding of the other sentence on the dependency parse tree. We evaluate the proposed attentive encoder on three tasks: semantic similarity, paraphrase identification and true-false question selection. Experimental results show that our encoder outperforms all baselines and achieves state-of-the-art results on two tasks.