EMNLP 2023 论文列表

Pretrained language models (PLMs), especially large language models (LLMs) demonstrate impressive capabilities in open-ended text generation. While our statistical results show that LLMs often suffer from over-concentrated information, where the generated texts overly focus on the given prompt and fail to provide sufficient background and detailed information as humans do. To address this issue, we propose a dynamic knowledge-guided informative open-ended text generation approach, that utilizes a knowledge graph to help the model generate more contextually related entities and detailed facts. Specifically, we first employ a local knowledge filter to extract relevant knowledge from the comprehensive knowledge graph for a given topic sentence. Then we introduce a dynamic knowledge selector to predict the entity to be mentioned in the subsequent sentence. Finally, we utilize a knowledge-enhanced text generator to produce a more informative output. To evaluate the effectiveness of our approach, we evaluate the proposed approach in two scenarios: fine-tuning for small PLMs and prompt tuning for LLMs. Experimental results show that our approach could generate more informative texts than baselines.

Neural document rerankers are extremely effective in terms of accuracy. However, the best models require dedicated hardware for serving, which is costly and often not feasible. To avoid this servingtime requirement, we present a method of capturing up to 86% of the gains of a Transformer cross-attention model with a lexicalized scoring function that only requires 10−6% of the Transformer’s FLOPs per document and can be served using commodity CPUs. When combined with a BM25 retriever, this approach matches the quality of a state-of-the art dual encoder retriever, that still requires an accelerator for query encoding. We introduce nail (Non-Autoregressive Indexing with Language models) as a model architecture that is compatible with recent encoder-decoder and decoder-only large language models, such as T5, GPT-3 and PaLM. This model architecture can leverage existing pre-trained checkpoints and can be fine-tuned for efficiently constructing document representations that do not require neural processing of queries.

Modular neural networks without additional training have recently been shown to surpass end-to-end neural networks on challenging vision–language tasks. The latest such methods simultaneously introduce LLM-based code generation to build programs and a number of skill-specific, task-oriented modules to execute them. In this paper, we focus on ViperGPT and ask where its additional performance comes from and how much is due to the (state-of-art, end-to-end) BLIP-2 model it subsumes vs. additional symbolic components. To do so, we conduct a controlled study (comparing end-to-end, modular, and prompting-based methods across several VQA benchmarks). We find that ViperGPT's reported gains over BLIP-2 can be attributed to its selection of task-specific modules, and when we run ViperGPT using a more task-agnostic selection of modules, these gains go away. ViperGPT retains much of its performance if we make prominent alterations to its selection of modules: e.g. removing or retaining only BLIP-2. We also compare ViperGPT against a prompting-based decomposition strategy and find that, on some benchmarks, modular approaches significantly benefit by representing subtasks with natural language, instead of code. Our code is fully available at https://github.com/brown-palm/visual-question-decomposition.

Conventional approaches to relation extraction can only recognize predefined relation types. In the real world, new or out-of-scope relation types may keep challenging the deployed models. In this paper, we formalize such a challenging problem as Novel Relation Detection (NRD), which aims to discover potential new relation types based on training samples of known relations. To this end, we construct two NRD datasets and exhaustively investigate a variety of out-of-scope detection methods. We further propose an effective NRD method that utilizes multi-strategy self-supervised learning to handle the problem of shallow semantic similarity in the NRD task. Experimental results demonstrate the effectiveness of our method, which significantly outperforms previous state-of-the-art methods on both datasets.

TTransformer models have demonstrated remarkable performance in neural machine translation (NMT). However, their vulnerability to noisy input poses a significant challenge in practical implementation, where generating clean output from noisy input is crucial. The MTNT dataset is widely used as a benchmark for evaluating the robustness of NMT models against noisy input. Nevertheless, its utility is limited due to the presence of noise in both the source and target sentences. To address this limitation, we focus on cleaning the noise from the target sentences in MTNT, making it more suitable as a benchmark for noise evaluation. Leveraging the capabilities of large language models (LLMs), we observe their impressive abilities in noise removal. For example, they can remove emojis while considering their semantic meaning. Additionally, we show that LLM can effectively rephrase slang, jargon, and profanities. The resulting datasets, called C-MTNT, exhibit significantly less noise in the target sentences while preserving the semantic integrity of the original sentences. Our human and GPT-4 evaluations also lead to a consistent conclusion that LLM performs well on this task. Lastly, experiments on C-MTNT showcased its effectiveness in evaluating the robustness of NMT models, highlighting the potential of advanced language models for data cleaning and emphasizing C-MTNT as a valuable resource.

While most task-oriented dialogues assume conversations between the agent and one user at a time, dialogue systems are increasingly expected to communicate with multiple users simultaneously who make decisions collaboratively. To facilitate development of such systems, we release the Multi-User MultiWOZ dataset: task-oriented dialogues among two users and one agent. To collect this dataset, each user utterance from MultiWOZ 2.2 was replaced with a small chat between two users that is semantically and pragmatically consistent with the original user utterance, thus resulting in the same dialogue state and system response. These dialogues reflect interesting dynamics of collaborative decision-making in task-oriented scenarios, e.g., social chatter and deliberation. Supported by this data, we propose the novel task of multi-user contextual query rewriting: to rewrite a task-oriented chat between two users as a concise task-oriented query that retains only task-relevant information and that is directly consumable by the dialogue system. We demonstrate that in multi-user dialogues, using predicted rewrites substantially improves dialogue state tracking without modifying existing dialogue systems that are trained for single-user dialogues. Further, this method surpasses training a medium-sized model directly on multi-user dialogues and generalizes to unseen domains.

Recent work has shown the promise of learning with human feedback paradigms to produce human-determined high-quality text. Existing works use human feedback to train large language models (LLMs) in general domain abstractive summarization and have obtained summary quality exceeding traditional likelihood training. In this paper, we focus on a less explored form of human feedback -- Human Edits. We propose Sequence Alignment (un)Likelihood Training (SALT), a novel technique to use both the human-edited and model-generated data together in the training loop. In addition, we demonstrate simulating Human Edits with ground truth summaries coming from existing training data -- Imitation edits, along with the model-generated summaries obtained after the training, to reduce the need for expensive human-edit data. In our experiments, we extend human feedback exploration from general domain summarization to medical domain summarization. Our results demonstrate the effectiveness of SALT in improving the summary quality with Human and Imitation Edits. Through additional experiments, we show that SALT outperforms the conventional RLHF method (designed for human preferences) -- DPO, when applied to human-edit data. We hope the evidence in our paper prompts researchers to explore, collect, and better use different human feedback approaches scalably.

Extractive summarization is a crucial task in natural language processing that aims to condense long documents into shorter versions by directly extracting sentences. The recent introduction of large language models has attracted significant interest in the NLP community due to its remarkable performance on a wide range of downstream tasks. This paper first presents a thorough evaluation of ChatGPT's performance on extractive summarization and compares it with traditional fine-tuning methods on various benchmark datasets. Our experimental analysis reveals that ChatGPT exhibits inferior extractive summarization performance in terms of ROUGE scores compared to existing supervised systems, while achieving higher performance based on LLM-based evaluation metrics. In addition, we explore the effectiveness of in-context learning and chain-of-thought reasoning for enhancing its performance. Furthermore, we find that applying an extract-then-generate pipeline with ChatGPT yields significant performance improvements over abstractive baselines in terms of summary faithfulness. These observations highlight potential directions for enhancing ChatGPT's capabilities in faithful summarization using two-stage approaches.

Prompt engineering, as an efficient and effective way to leverage Large Language Models (LLM), has drawn a lot of attention from the research community. The existing research primarily emphasizes the importance of adapting prompts to specific tasks, rather than specific LLMs. However, a good prompt is not solely defined by its wording, but also binds to the nature of the LLM in question. In this work, we first quantitatively demonstrate that different prompts should be adapted to different LLMs to enhance their capabilities across various downstream tasks in NLP. Then we novelly propose a model-adaptive prompt optimizer (MAPO) method that optimizes the original prompts for each specific LLM in downstream tasks. Extensive experiments indicate that the proposed method can effectively refine prompts for an LLM, leading to significant improvements over various downstream tasks.

We illustrate how a calibrated model can help balance common trade-offs in task-oriented parsing. In a simulated annotator-in-the-loop experiment, we show that well-calibrated confidence scores allow us to balance cost with annotator load, improving accuracy with a small number of interactions. We then examine how confidence scores can help optimize the trade-off between usability and safety. We show that confidence-based thresholding can substantially reduce the number of incorrect low-confidence programs executed; however, this comes at a cost to usability. We propose the DidYouMean system which better balances usability and safety by rephrasing low-confidence inputs.

Recent advances in large language models (LLMs), such as ChatGPT, have showcased remarkable zero-shot performance across various NLP tasks. However, the potential of LLMs in personality detection, which involves identifying an individual's personality from their written texts, remains largely unexplored. Drawing inspiration from Psychological Questionnaires, which are carefully designed by psychologists to evaluate individual personality traits through a series of targeted items, we argue that these items can be regarded as a collection of well-structured chain-of-thought (CoT) processes. By incorporating these processes, LLMs can enhance their capabilities to make more reasonable inferences on personality from textual input. In light of this, we propose a novel personality detection method, called PsyCoT, which mimics the way individuals complete psychological questionnaires in a multi-turn dialogue manner. In particular, we employ a LLM as an AI assistant with a specialization in text analysis. We prompt the assistant to rate individual items at each turn and leverage the historical rating results to derive a conclusive personality preference. Our experiments demonstrate that PsyCoT significantly improves the performance and robustness of GPT-3.5 in personality detection, achieving an average F1 score improvement of 4.23/10.63 points on two benchmark datasets compared to the standard prompting method. Our code is available at \url{https://github.com/TaoYang225/PsyCoT.

Instruction tuned large language models (LLMs), such as ChatGPT, demonstrate remarkable performance in a wide range of tasks. Despite numerous recent studies that examine the performance of instruction-tuned LLMs on various NLP benchmarks, there remains a lack of comprehensive investigation into their ability to understand cross-lingual sociopragmatic meaning (SM), i.e., meaning embedded within social and interactive contexts. This deficiency arises partly from SM not being adequately represented in any of the existing benchmarks. To address this gap, we present SPARROW, an extensive multilingual benchmark specifically designed for SM understanding. SPARROW comprises 169 datasets covering 13 task types across six primary categories (e.g., anti-social language detection, emotion recognition). SPARROW datasets encompass 64 different languages originating from 12 language families representing 16 writing scripts. We evaluate the performance of various multilingual pretrained language models (e.g., mT5) and instruction-tuned LLMs (e.g., BLOOMZ, ChatGPT) on SPARROW through fine-tuning, zero-shot, and/or few-shot learning. Our comprehensive analysis reveals that existing open-source instruction tuned LLMs still struggle to understand SM across various languages, performing close to a random baseline in some cases. We also find that although ChatGPT outperforms many LLMs, it still falls behind task-specific finetuned models with a gap of 12.19 SPARROW score. Our benchmark is available at: https://github.com/UBC-NLP/SPARROW

To explore how humans can best leverage LLMs for writing and how interacting with these models affects feelings of ownership and trust in the writing process, we compared common human-AI interaction types (e.g., guiding system, selecting from system outputs, post-editing outputs) in the context of LLM-assisted news headline generation. While LLMs alone can generate satisfactory news headlines, on average, human control is needed to fix undesirable model outputs. Of the interaction methods, guiding and selecting model output added the most benefit with the lowest cost (in time and effort). Further, AI assistance did not harm participants’ perception of control compared to freeform editing.

Large Language Models (LLMs) trained with self-supervision on vast corpora of web text fit to the social biases of that text. Without intervention, these social biases persist in the model's predictions in downstream tasks, leading to representational harm. Many strategies have been proposed to mitigate the effects of inappropriate social biases learned during pretraining. Simultaneously, methods for model compression have become increasingly popular to reduce the computational burden of LLMs. Despite the popularity and need for both approaches, little work has been done to explore the interplay between these two. We perform a carefully controlled study of the impact of model compression via quantization and knowledge distillation on measures of social bias in LLMs. Longer pretraining and larger models led to higher social bias, and quantization showed a regularizer effect with its best trade-off around 20\% of the original pretraining time.

The ability to automatically generate accurate protocols for scientific experiments would represent a major step towards the automation of science. Large Language Models (LLMs) have impressive capabilities on a wide range of tasks, such as question answering and the generation of coherent text and code. However, LLMs can struggle with multi-step problems and long-term planning, which are crucial for designing scientific experiments. Moreover, evaluation of the accuracy of scientific protocols is challenging, because experiments can be described correctly in many different ways, require expert knowledge to evaluate, and cannot usually be executed automatically. Here we present an automatic evaluation framework for the task of planning experimental protocols, and we introduce BioProt: a dataset of biology protocols with corresponding pseudocode representations. To measure performance on generating scientific protocols, we use an LLM to convert a natural language protocol into pseudocode, and then evaluate an LLM's ability to reconstruct the pseudocode from a high-level description and a list of admissible pseudocode functions. We evaluate GPT-3 and GPT-4 on this task and explore their robustness. We externally validate the utility of pseudocode representations of text by generating accurate novel protocols using retrieved pseudocode, and we run a generated protocol successfully in our biological laboratory. Our framework is extensible to the evaluation and improvement of language model

Recognizing entities in texts is a central need in many information-seeking scenarios, and indeed, Named Entity Recognition (NER) is arguably one of the most successful examples of a widely adopted NLP task and corresponding NLP technology. Recent advances in large language models (LLMs) appear to provide effective solutions (also) for NER tasks that were traditionally handled with dedicated models, often matching or surpassing the abilities of the dedicated models. Should NER be considered a solved problem? We argue to the contrary: the capabilities provided by LLMs are not the end of NER research, but rather an exciting beginning. They allow taking NER to the next level, tackling increasingly more useful, and increasingly more challenging, variants. We present three variants of the NER task, together with a dataset to support them. The first is a move towards more fine-grained---and intersectional---entity types. The second is a move towards zero-shot recognition and extraction of these fine-grained types based on entity-type labels. The third, and most challenging, is the move from the recognition setup to a novel retrieval setup, where the query is a zero-shot entity type, and the expected result is all the sentences from a large, pre-indexed corpus that contain entities of these types, and their corresponding spans. We show that all of these are far from being solved. We provide a large, silver-annotated corpus of 4 million paragraphs covering 500 entity types, to facilitate research towards all of these three goals.

Chain-of-thought (CoT) is capable of eliciting models to explicitly generate reasoning paths, thus promoting reasoning accuracy and attracting increasing attention. Specifically, zero-shot CoT achieves remarkable improvements in a wide range of reasoning tasks by simply instructing the LLM with the prompt "Let's think step by step!". Despite the success of zero-shot CoT, the existing zero-shot prompting techniques remain limited to a single language, making it challenging to generalize to other languages and hindering global development. In this work, we introduce cross-lingual prompting (CLP), aiming to improve zero-shot CoT reasoning across languages. Specifically, CLP consists of two main components: (1) cross-lingual alignment prompting and (2) task-specific solver prompting. The cross-lingual alignment prompting is responsible for aligning representations across different languages, whereas the task-specific solver prompting is used to generate the final chain of thoughts and results for the reasoning task. In addition, we further introduce cross-lingual self-consistent prompting (CLSP) to ensemble different reasoning paths across languages. Our experimental evaluations on several benchmarks demonstrate that CLP and CLSP significantly outperform the existing prompting methods and achieve state-of-the-art performance. We hope this work will inspire further breakthroughs in cross-lingual CoT.

Large language models (LLMs) excel in many tasks in NLP and beyond, but most open models have very limited coverage of smaller languages and LLM work tends to focus on languages where nearly unlimited data is available for pretraining. In this work, we study the challenges of creating LLMs for Finnish, a language spoken by less than 0.1\% of the world population. We compile an extensive dataset of Finnish combining web crawls, news, social media and eBooks. We pursue two approaches to pretrain models: 1) we train seven monolingual models from scratch (186M to 13B parameters) dubbed FinGPT, 2) we continue the pretraining of the multilingual BLOOM model on a mix of its original training data and Finnish, resulting in a 176 billion parameter model we call BLUUMI. For model evaluation, we introduce FIN-bench, a version of BIG-bench with Finnish tasks. We also assess other model qualities such as toxicity and bias. Our models and tools are openly available at \url{https://turkunlp.org/gpt3-finnish}.

This paper presents FlowSUM, a normalizing flows-based variational encoder-decoder framework for Transformer-based summarization. Our approach tackles two primary challenges in variational summarization: insufficient semantic information in latent representations and posterior collapse during training. To address these challenges, we employ normalizing flows to enable flexible latent posterior modeling, and we propose a controlled alternate aggressive training (CAAT) strategy with an improved gate mechanism. Experimental results show that FlowSUM significantly enhances the quality of generated summaries and unleashes the potential for knowledge distillation with minimal impact on inference time. Furthermore, we investigate the issue of posterior collapse in normalizing flows and analyze how the summary quality is affected by the training strategy, gate initialization, and the type and number of normalizing flows used, offering valuable insights for future research.