论文检索

Joan Santoso, Patrick Sutanto, Billy Cahyadi, Esther Setiawan

Named Entity Recognition (NER) is an important task, but to achieve great performance, it is usually necessary to collect a large amount of labeled data, incurring high costs. In this paper, we propose using open-source Large Language Models (LLM) to generate NER data with only a few labeled examples, reducing the cost of human annotations. Our proposed method is very simple and can perform well using only a few labeled data points. Experimental results on diverse low-resource NER datasets show that our proposed data generation method can significantly improve the baseline. Additionally, our method can be used to augment datasets with class-imbalance problems and consistently improves model performance on macro-F1 metrics.

Ran Liu, Ming Liu, Min Yu, He Zhang, Jianguo Jiang, Gang Li, Weiqing Huang

With the popularity of large language models (LLMs) and their ability to handle longer input documents, there is a growing need for high-quality long document summarization datasets. Although many models already support 16k input, current lengths of summarization datasets are inadequate, and salient information is not evenly distributed. To bridge these gaps, we collect a new summarization dataset called SumSurvey, consisting of more than 18k scientific survey papers. With an average document length exceeding 12k and a quarter exceeding 16k, as well as the uniformity metric outperforming current mainstream long document summarization datasets, SumSurvey brings new challenges and expectations to both fine-tuned models and LLMs. The informativeness of summaries and the models supporting the evaluation of long document summarization warrant further attention. Automatic and human evaluation results on this abstractive dataset confirm this view. Our dataset and code are available at https://github.com/Oswald1997/SumSurvey.

Wen Huang, Hongbin Liu, Minxin Guo, Neil Gong

Visual hallucination (VH) means that a multi-modal LLM (MLLM) imagines incorrect details about an image in visual question answering. Existing studies find VH instances only in existing image datasets, which results in biased understanding of MLLMs’ performance under VH due to limited diversity of such VH instances. In this work, we propose a tool called VHTest to generate a diverse set of VH instances. Specifically, VHTest finds some initial VH instances in existing image datasets (e.g., COCO), generates a text description for each VH mode, and uses a text-to-image generative model (e.g., DALL-E-3) to generate VH images based on the text descriptions. We collect a benchmark dataset with 1,200 VH instances in 8 VH modes using VHTest. We find that existing MLLMs such as GPT-4, LLaVA-1.5, and MiniGPT-v2 hallucinate for a large fraction of the instances in our benchmark. Moreover, we find that fine-tuning an MLLM using our benchmark dataset reduces its likelihood to hallucinate without sacrificing its performance on other benchmarks. Our benchmarks are publicly available: https://github.com/wenhuang2000/VHTest.

Frank Wildenburg, Michael Hanna, Sandro Pezzelle

In everyday language use, speakers frequently utter and interpret sentences that are semantically underspecified, namely, whose content is insufficient to fully convey their message or interpret them univocally. For example, to interpret the underspecified sentence “Don’t spend too much”, which leaves implicit what (not) to spend, additional linguistic context or outside knowledge is needed. In this work, we propose a novel Dataset of semantically Underspecified Sentences grouped by Type (DUST) and use it to study whether pre-trained language models (LMs) correctly identify and interpret underspecified sentences. We find that newer LMs are reasonably able to identify underspecified sentences when explicitly prompted. However, interpreting them correctly is much harder for any LMs. Our experiments show that when interpreting underspecified sentences, LMs exhibit little uncertainty, contrary to what theoretical accounts of underspecification would predict. Overall, our study reveals limitations in current models’ processing of sentence semantics and highlights the importance of using naturalistic data and communicative scenarios when evaluating LMs’ language capabilities.

Anton Schäfer, Thomas Hofmann, Imanol Schlag, Tiago Pimentel

Tokenisation is a core part of language models (LMs). It involves splitting a character sequence into subwords which are assigned random indices before being served to the LM. However, this process—while typically lossless—may lead to less efficient LM training, because it removes character-level information, thereby making it more difficult to generalise across similar subwords, such as *now* and *Now*. We refer to such subwords as **near duplicates**. In this paper, we study the impact of near duplicate subwords on LM training efficiency. First, we design an experiment that gives us an upper bound to how much we should expect a model to improve if we could perfectly generalise across near duplicates. We do this, by duplicating each token in our LM’s vocabulary, creating perfectly equivalent classes of subwords. Experimentally, we find that LMs need roughly 17% more data when trained in a fully duplicated setting. Second, we investigate the impact of naturally occurring near duplicates on LMs. Here, we see that deduplicating them considerably hurts LM performance; but that this loss in performance can be easily mitigated.

Erxin Yu, Jing Li, Chunpu Xu

A quote tweet enables users to share others’ content while adding their own commentary. In order to enhance public engagement through quote tweets, we investigate the task of generating popular quote tweets. This task aims to produce quote tweets that garner higher popularity, as indicated by increased likes, replies, and retweets. Despite the impressive language generation capabilities of large language models (LLMs), there has been limited research on how LLMs can effectively learn the popularity of text to better engage the public. Therefore, we introduce a novel approach called Response-augmented Popularity-Aligned Language Model (RePALM), which aligns language generation with popularity by leveraging insights from augmented auto-responses provided by readers. We utilize the Proximal Policy Optimization framework with a dual-reward mechanism to jointly optimize for the popularity of the quote tweet and its consistency with the auto-responses. In our experiments, we collected two datasets consisting of quote tweets containing external links and those referencing others’ tweets. Extensive results demonstrate the superiority of RePALM over advanced language models that do not incorporate response augmentation.

Yinuo Jiang, Xiang Zhuang, Keyan Ding, Qiang Zhang, Huajun Chen

The development of Large Language Models (LLMs) has greatly advanced the field of drug discovery, with the belief that natural language can enhance human control over molecule design. However, the scarcity of high-quality labeled data remains a challenge for cross text-molecule learning. Existing datasets are limited due to the difficulty of collecting precise molecule-description pairs. Although recent efforts have utilized pseudo data generated by LLMs for augmentation, the lack of specialized chemistry knowledge of LLMs and the absence of an effective high quality data selector may introduce noise into the annotations, compromising the models’ robustness. To address these challenges, this paper introduces a novel framework that interweaves model fine-tuning and data augmentation to overcome the scarcity of high-quality data. The proposed approach involves an iterative procedure where the model plays dual roles in annotating unlabeled data and sampling a subset of high-quality data until convergence is achieved, enhancing the model’s understanding and adaptability. Additionally, a new dataset called SAPubChem-41 is presented, which comprises meticulously curated high-quality parallel molecule-description pairs designed specifically for fine-tuning purposes. This research provides an important contribution to the field by addressing the need for high-quality datasets and presenting an effective framework for cross text-molecule learning.

Gili Lior, Yoav Goldberg, Gabriel Stanovsky

Document collections of various domains, e.g., legal, medical, or financial, often share some underlying collection-wide structure, which captures information that can aid both human users and structure-aware models.We propose to identify the typical structure of document within a collection, which requires to capture recurring topics across the collection, while abstracting over arbitrary header paraphrases, and ground each topic to respective document locations. These requirements pose several challenges: headers that mark recurring topics frequently differ in phrasing, certain section headers are unique to individual documents and do not reflect the typical structure, and the order of topics can vary between documents. Subsequently, we develop an unsupervised graph-based method which leverages both inter- and intra-document similarities, to extract the underlying collection-wide structure. Our evaluations on three diverse domains in both English and Hebrew indicate that our method extracts meaningful collection-wide structure, and we hope that future work will leverage our method for multi-document applications and structure-aware models.

Renzhi Wang, Piji Li

Parameter-Efficient Fine-Tuning (PEFT) methods enable efficient adaptation of Large Language Models (LLMs) to various downstream applications. However, the effectiveness of the PEFT diminishes notably when downstream tasks require accurate learning of specific knowledge. In this paper, we adopt a semantic perspective to investigate this phenomenon, uncovering the reasons behind PEFT’s limitations in knowledge learning task. Our findings reveals that: (1) PEFT presents a notable risk of pushing the model away from the intended knowledge target; (2) multiple knowledge interfere with each other, and such interference suppresses the learning and expression of knowledge features. Based on these insights, we introduce a data filtering strategy to exclude data that is detrimental to knowledge learning and a re-weighted learning strategy to make the model attentive to semantic distance during knowledge learning. Experimental results demonstrate the effectiveness of the proposed method on open-source large language model, further validate the semantic challenge in PEFT, thus paving the way for future research.

Dennis Ulmer, Elman Mansimov, Kaixiang Lin, Lijia Sun, Xibin Gao, Yi Zhang

Large language models (LLMs) are powerful dialogue agents, but specializing them towards fulfilling a specific function can be challenging. Instructing tuning, i.e. tuning models on instruction and sample responses generated by humans (Ouyang et al., 2022), has proven as an effective method to do so, yet requires a number of data samples that a) might not be available or b) costly to generate. Furthermore, this cost increases when the goal is to make the LLM follow a specific workflow within a dialogue instead of single instructions. Inspired by the self-play technique in reinforcement learning and the use of LLMs to simulate human agents, we propose a more effective method for data collection through LLMs engaging in a conversation in various roles. This approach generates a training data via “self-talk” of LLMs that can be refined and utilized for supervised fine-tuning. We introduce an automated way to measure the (partial) success of a dialogue. This metric is used to filter the generated conversational data that is fed back in LLM for training. Based on our automated and human evaluations of conversation quality, we demonstrate that such self-talk data improves results. In addition, we examine the various characteristics that showcase the quality of generated dialogues and how they can be connected to their potential utility as training data.

Chen Xu, Jie Wang, Xiaoqian Liu, Qian Dong, Chunliang Zhang, Tong Xiao, JingBo Zhu, Dapeng Man, Wu Yang

Speech-to-text (S2T) generation systems frequently face challenges in low-resource scenarios, primarily due to the lack of extensive labeled datasets. One emerging solution is constructing virtual training samples by interpolating inputs and labels, which has notably enhanced system generalization in other domains. Despite its potential, this technique’s application in S2T tasks has remained under-explored. In this paper, we delve into the utility of interpolation augmentation, guided by several pivotal questions. Our findings reveal that employing an appropriate strategy in interpolation augmentation significantly enhances performance across diverse tasks, architectures, and data scales, offering a promising avenue for more robust S2T systems in resource-constrained settings.

Wanlong Liu, Li Zhou, DingYi Zeng, Yichen Xiao, Shaohuan Cheng, Chen Zhang, Grandee Lee, Malu Zhang, Wenyu Chen

Recent mainstream event argument extraction methods process each event in isolation, resulting in inefficient inference and ignoring the correlations among multiple events. To address these limitations, here we propose a multiple-event argument extraction model DEEIA (Dependency-guided Encoding and Event-specific Information Aggregation), capable of extracting arguments from all events within a document simultaneously. The proposed DEEIA model employs a multi-event prompt mechanism, comprising DE and EIA modules. The DE module is designed to improve the correlation between prompts and their corresponding event contexts, whereas the EIA module provides event-specific information to improve contextual understanding. Extensive experiments show that our method achieves new state-of-the-art performance on four public datasets (RAMS, WikiEvents, MLEE, and ACE05), while significantly saving the inference time compared to the baselines. Further analyses demonstrate the effectiveness of the proposed modules.

Zhexuan Wang, Shudong Liu, Xuebo Liu, Miao Zhang, Derek Wong, Min Zhang

kNN-MT has utilized neighborhood knowledge for auxiliary decoding, significantly improving translation performance. Subsequently, kNN-KD transitions the use of neighborhood knowledge from the decoding phase to the training phase, to address the temporal and spatial inefficiencies inherent in kNN-MT. However, kNN-KD transfers all the kNN knowledge arbitrarily, which has the potential to restrict the learning of student models. In this paper, we propose a novel domain-aware kNN-KD method, which filters out domain-relevant neighborhood knowledge for learning in the distillation process. Notably, this entire process exclusively utilizes the neighborhood knowledge of the original model, eliminating the need for establishing any additional datastores. Experiments on four domain translation tasks demonstrate that our method achieves state-of-the-art performance, realizing an average gain of 1.55 COMET and 1.42 BLEU scores, by further enhancing the translation of rare words. Source code can be accessed at https://github.com/wangzx1219/Dk-KD.

Gabriele Picco, Leopold Fuchs, Marcos Martínez Galindo, Alberto Purpura, Vanessa López, Hoang Thanh Lam

Zero-shot entity and relation classification models leverage available external information of unseen classes – e.g., textual descriptions – to annotate input text data. Thanks to the minimum data requirement, Zero-Shot Learning (ZSL) methods have high value in practice, especially in applications where labeled data is scarce. Even though recent research in ZSL has demonstrated significant results, our analysis reveals that those methods are sensitive to provided textual descriptions of entities (or relations). Even a minor modification of descriptions can lead to a change in the decision boundary between entity (or relation) classes. In this paper, we formally define the problem of identifying effective descriptions for zero shot inference. We propose a strategy for generating variations of an initial description, a heuristic for ranking them and an ensemble method capable of boosting the predictions of zero-shot models through description enhancement. Empirical results on four different entity and relation classification datasets show that our proposed method outperform existing approaches and achieve new SOTA results on these datasets under the ZSL settings. The source code of the proposed solutions and the evaluation framework are open-sourced.

Luca Ragazzi, Paolo Italiani, Gianluca Moro, Mattia Panni

Scientific document summarization aims to condense complex and long articles in both technical and plain-language terms to facilitate the accessibility and dissemination of scientific findings. Existing datasets suffer from a deficiency in source heterogeneity, as their data predominantly stem from a single common resource, hindering effective model training and generalizability. First, we introduce SciLay, a novel dataset that includes documents from multiple natural science journals with expert-authored technical and lay summaries. Second, we propose PrunePert, a new transformer-based model that incorporates a differentiable perturbed top-k encoder layer to prune irrelevant tokens in end-to-end learning. Experimental results show that our model achieves a nearly 2x speed-up compared to a state-of-the-art linear transformer, remaining comparable in effectiveness. Additional examinations underscore the importance of employing a training dataset that includes different sources to enhance the generalizability of the models. Code is available at https://github.com/disi-unibo-nlp/sci-lay.

Everlyn Asiko Chimoto, Jay Gala, Orevaoghene Ahia, Julia Kreutzer, Bruce A. Bassett, Sara Hooker

Neural Machine Translation models are extremely data and compute-hungry. However, not all datapoints contribute equally to model training and generalization. Data pruning to remove the low-value data points has the benefit of drastically reducing the compute budget without significantdrop in model performance. In this paper, we propose a new data pruning technique: CheckpointsAcross Time (CAT ), that leverages early model training dynamics to identify the most relevantdata points for model performance. We benchmark CAT against several data pruning techniquesincluding COMET-QE, LASER and LaBSE. We find that CAT outperforms the benchmarks onIndo-European languages on multiple test sets. When applied to English-German, English-Frenchand English-Swahili translation tasks, CAT achieves comparable performance to using the fulldataset, while pruning up to 50% of training data. We inspect the data points that CAT selectsand find that it tends to favour longer sentences and sentences with unique or rare words.

Yang Zhao, Li Du, Xiao Ding, Kai Xiong, Zhouhao Sun, Shi Jun, Ting Liu, Bing Qin

Through pretraining on a corpus with various sources, Large Language Models (LLMs) have gained impressive performance. However, the impact of each component of the pretraining corpus remains opaque. As a result, the organization of the pretraining corpus is still empirical and may deviate from the optimal. To address this issue, we systematically analyze the impact of 48 datasets from 5 major categories of pretraining data of LLMs and measure their impacts on LLMs using benchmarks about nine major categories of model capabilities. Our analyses provide empirical results about the contribution of multiple corpora on the performances of LLMs, along with their joint impact patterns, including complementary, orthogonal, and correlational relationships. We also identify a set of “high-impact data” such as Books that is significantly related to a set of model capabilities. These findings provide insights into the organization of data to support more efficient pretraining of LLMs.

Ekaterina Fadeeva, Aleksandr Rubashevskii, Artem Shelmanov, Sergey Petrakov, Haonan Li, Hamdy Mubarak, Evgenii Tsymbalov, Gleb Kuzmin, Alexander Panchenko, Timothy Baldwin 等

Large language models (LLMs) are notorious for hallucinating, i.e., producing erroneous claims in their output. Such hallucinations can be dangerous, as occasional factual inaccuracies in the generated text might be obscured by the rest of the output being generally factually correct, making it extremely hard for the users to spot them. Current services that leverage LLMs usually do not provide any means for detecting unreliable generations. Here, we aim to bridge this gap. In particular, we propose a novel fact-checking and hallucination detection pipeline based on token-level uncertainty quantification. Uncertainty scores leverage information encapsulated in the output of a neural network or its layers to detect unreliable predictions, and we show that they can be used to fact-check the atomic claims in the LLM output. Moreover, we present a novel token-level uncertainty quantification method that removes the impact of uncertainty about what claim to generate on the current step and what surface form to use. Our method Claim Conditioned Probability (CCP) measures only the uncertainty of a particular claim value expressed by the model. Experiments on the task of biography generation demonstrate strong improvements for CCP compared to the baselines for seven different LLMs and four languages. Human evaluation reveals that the fact-checking pipeline based on uncertainty quantification is competitive with a fact-checking tool that leverages external knowledge.

Zehui Chen, Kuikun Liu, Qiuchen Wang, Wenwei Zhang, Jiangning Liu, Dahua Lin, Kai Chen, Feng Zhao

Open-sourced Large Language Models (LLMs) have achieved great success in various NLP tasks, however, they are still far inferior to API-based models when acting as agents. How to integrate agent ability into general LLMs becomes a crucial and urgent problem.This paper first delivers three key observations: (1) the current agent training corpus is entangled with both formats following and agent reasoning, which significantly shifts from the distribution of its pre-training data; (2) LLMs exhibit different learning speeds on the capabilities required by agent tasks; and (3) current approaches have side-effects when improving agent abilities by introducing hallucinations. Based on the above findings, we propose Agent-FLAN to effectively Fine-tune LANguage models for Agents.Through careful decomposition and redesign of the training corpus, Agent-FLAN enables Llama2-7B to outperform prior best works by 3.5% across various agent evaluation datasets. With comprehensively constructed negative samples, Agent-FLAN greatly alleviates the hallucination issues based on our established evaluation benchmark. Besides, it consistently improves the agent capability of LLMs when scaling model sizes while slightly enhancing the general capability of LLMs. The code and models are available at https://github.com/InternLM/Agent-FLAN.

Huanhuan Ma, Weizhi Xu, Yifan Wei, Liuji Chen, Liang Wang, Qiang Liu, Shu Wu, Liang Wang

Fact verification aims to automatically probe the veracity of a claim based on several pieces of evidence. Existing works are always engaging in accuracy improvement, let alone explainability, a critical capability of fact verification systems.Constructing an explainable fact verification system in a complex multi-hop scenario is consistently impeded by the absence of a relevant, high-quality dataset. Previous datasets either suffer from excessive simplification or fail to incorporate essential considerations for explainability. To address this, we present EX-FEVER, a pioneering dataset for multi-hop explainable fact verification. With over 60,000 claims involving 2-hop and 3-hop reasoning, each is created by summarizing and modifying information from hyperlinked Wikipedia documents. Each instance is accompanied by a veracity label and an explanation that outlines the reasoning path supporting the veracity classification. Additionally, we demonstrate a novel baseline system on our EX-FEVER dataset, showcasing document retrieval, explanation generation, and claim verification, and validate the significance of our dataset. Furthermore, we highlight the potential of utilizing Large Language Models in the fact verification task. We hope our dataset could make a significant contribution by providing ample opportunities to explore the integration of natural language explanations in the domain of fact verification.