Long-context modeling capabilities of Large Language Models (LLMs) have garnered widespread attention, leading to the emergence of LLMs with ultra-context windows. Meanwhile, benchmarks for evaluating long-context language models are gradually catching up. However, existing benchmarks employ irrelevant noise texts to artificially extend the length of test cases, diverging from the real-world scenarios of long-context applications. To bridge this gap, we propose a novel long-context benchmark, Loong, aligning with realistic scenarios through extended multi-document question answering (QA). Unlike typical document QA, in Loong’s test cases, each document is relevant to the final answer, ignoring any document will lead to the failure of the answer. Furthermore, Loong introduces four types of tasks with a range of context lengths: Spotlight Locating, Comparison, Clustering, and Chain of Reasoning, to facilitate a more realistic and comprehensive evaluation of long-context understanding. Extensive experiments indicate that existing long-context language models still exhibit considerable potential for enhancement. Retrieval augmented generation (RAG) achieves poor performance, demonstrating that Loong can reliably assess the model’s long-context modeling capabilities.

Spoken Named Entity Recognition (NER) aims to extract entities from speech. The extracted entities can help voice assistants better understand user’s questions and instructions. However, current Chinese Spoken NER datasets are laboratory-controlled data that are collected by reading existing texts in quiet environments, rather than natural spoken data, and the texts used for reading are also limited in topics. These limitations obstruct the development of Spoken NER in more natural and common real-world scenarios. To address this gap, we introduce a real-world Chinese Spoken NER dataset (RWCS-NER), encompassing open-domain daily conversations and task-oriented intelligent cockpit instructions. We compare several mainstream pipeline approaches on RWCS-NER. The results indicate that the current methods, affected by Automatic Speech Recognition (ASR) errors, do not perform satisfactorily in real settings. Aiming to enhance Spoken NER in real-world scenarios, we propose two approaches: self-training-asr and mapping then distilling (MDistilling). Experiments show that both approaches can achieve significant improvements, particularly MDistilling. Even compared with GPT4.0, MDistilling still reaches better results. We believe that our work will advance the field of Spoken NER in real-world settings.

Recently, tremendous strides have been made to align the generation of Large Language Models (LLMs) with human values to mitigate toxic or unhelpful content. Leveraging Reinforcement Learning from Human Feedback (RLHF) proves effective and is widely adopted by researchers. However, implementing RLHF is complex, and its sensitivity to hyperparameters renders achieving stable performance and scalability challenging. Furthermore, prevailing approaches to preference alignment primarily concentrate on pairwise comparisons, with limited exploration into multi-response scenarios, thereby overlooking the potential richness within the candidate pool. For the above reasons, we propose a new approach: Listwise Reward Enhancement for Preference Alignment (LIRE), a gradient-based reward optimization approach that incorporates the offline rewards of multiple responses into a streamlined listwise framework, thus eliminating the need for online sampling during training. LIRE is straightforward to implement, requiring minimal parameter tuning, and seamlessly aligns with the pairwise paradigm while naturally extending to multi-response scenarios. Moreover, we introduce a self-enhancement algorithm aimed at iteratively refining the reward during training. Our experiments demonstrate that LIRE consistently outperforms existing methods across several benchmarks on dialogue and summarization tasks, with good transferability to out-of-distribution data, assessed using proxy reward models and human annotators.

Text-based knowledge graph completion (KGC) methods utilize pre-trained language models for triple encoding and further fine-tune the model to achieve completion. Despite their excellent performance, they neglect the knowledge context in inferring process. Intuitively, knowledge contexts, which refer to the neighboring triples around the target triples, are important information for triple inferring, since they provide additional detailed information about the entities. To this end, we propose a novel framework named KnowC, which models the knowledge context as additional prompts with pre-trained language models for knowledge graph completion. Given the substantial number of neighbors typically associated with entities, along with the constrained input token capacity of language models, we further devise several strategies to sample the neighbors. We conduct extensive experiments on common datasets FB15k-237, WN18RR and Wikidata5M, experiments show that KnowC achieves state-of-the-art performance.

The advancement of large language models (LLMs) has enhanced the ability to generalize across a wide range of unseen natural language processing (NLP) tasks through instruction-following.Yet, their effectiveness often diminishes in low-resource languages like Chinese, exacerbated by biased evaluations from data leakage, casting doubt on their true generalizability to new linguistic territories. In response, we introduce the Chinese Instruction-Following Benchmark (**CIF-Bench**), designed to evaluate the zero-shot generalizability of LLMs to the Chinese language. CIF-Bench comprises 150 tasks and 15,000 input-output pairs, developed by native speakers to test complex reasoning and Chinese cultural nuances across 20 categories. To mitigate data contamination, we release only half of the dataset publicly, with the remainder kept private, and introduce diversified instructions to minimize score variance, totaling 45,000 data instances.Our evaluation of 28 selected LLMs reveals a noticeable performance gap, with the best model scoring only 52.9%, highlighting the limitations of LLMs in less familiar language and task contexts.This work not only uncovers the current limitations of LLMs in handling Chinese language tasks but also sets a new standard for future LLM generalizability research, pushing towards the development of more adaptable, culturally informed, and linguistically diverse models.

The instruction-following ability of large language models enables humans to interact with AI agents in a natural way. However, when required to generate responses of a specific length, large language models often struggle to meet users’ needs due to their inherent difficulty in accurately perceiving numerical constraints. To explore the ability of large language models to control the length of generated responses, we propose the Target Length Generation Task (TLG) and design two metrics, Precise Match (PM) and Flexible Match (FM) to evaluate the model’s performance in adhering to specified response lengths. Furthermore, we introduce a novel, model-agnostic approach called Ruler, which employs Meta Length Tokens (MLTs) to enhance the instruction-following ability of large language models under length-constrained instructions. Specifically, Ruler equips LLMs with the ability to generate responses of a specified length based on length constraints within the instructions. Moreover, Ruler can automatically generate appropriate MLT when length constraints are not explicitly provided, demonstrating excellent versatility and generalization. Comprehensive experiments show the effectiveness of Ruler across different LLMs on Target Length Generation Task, e.g., at All Level 27.97 average gain on PM, 29.57 average gain on FM. In addition, we conduct extensive ablation experiments to further substantiate the efficacy and generalization of Ruler. Our code and data is available on the internet.

The field of machine learning (ML) has gained widespread adoption, leading to significant demand for adapting ML to specific scenarios, which is yet expensive and non-trivial. The predominant approaches towards the automation of solving ML tasks (e.g., AutoML) are often time-consuming and hard to understand for human developers. In contrast, though human engineers have the incredible ability to understand tasks and reason about solutions, their experience and knowledge are often sparse and difficult to utilize by quantitative approaches. In this paper, we aim to bridge the gap between machine intelligence and human knowledge by introducing a novel framework MLCopilot, which leverages the state-of-the-art large language models to develop ML solutions for novel tasks. We showcase the possibility of extending the capability of LLMs to comprehend structured inputs and perform thorough reasoning for solving novel ML tasks. And we find that, after some dedicated design, the LLM can (i) observe from the existing experiences of ML tasks and (ii) reason effectively to deliver promising results for new tasks. The solution generated can be used directly to achieve high levels of competitiveness.

Contemporary practices in instruction tuning often hinge on enlarging data scaling without a clear strategy for ensuring data quality, inadvertently introducing noise that may compromise model performance. To address this challenge, we introduce Nuggets, a novel and efficient methodology that leverages one-shot learning to discern and select high-quality instruction data from extensive datasets. Nuggets assesses the potential of individual instruction examples to act as effective one-shot learning instances, thereby identifying those that can significantly improve performance across diverse tasks. Nuggets utilizes a scoring system based on the impact of candidate examples on the perplexity of a diverse anchor set, facilitating the selection of the most advantageous data for instruction tuning. Through rigorous evaluations on two benchmarks, namely MT-Bench and Alpaca-Eval, our study illustrates that instruction tuning with the top 1% of examples curated by Nuggets substantially outperforms conventional methods employing the entire dataset.

With the advancement of large language models (LLMs) and the expansion of their context windows, existing long-context benchmarks fall short in effectively evaluating the models’ comprehension and reasoning abilities in extended texts. Moreover, conventional benchmarks relying on F1 metrics often inaccurately score responses: they may undervalue correct answers that differ from the reference responses and overvalue incorrect ones that resemble the reference texts. In response to these limitations, we introduce Marathon, a novel evaluation benchmark that adopts a multiple-choice question format. It is specifically designed to overcome the constraints of previous benchmarks and provide a rapid, precise, and unbiased appraisal of the long-context comprehension skills of large language models. We conducted comprehensive evaluations on the Marathon benchmark with a range of state-of-the-art LLMs and assessed the effectiveness of various optimization strategies tailored for long-context generation. We anticipate that the Marathon benchmark and its associated leaderboard will enable a more precise and equitable evaluation of LLMs’ capabilities in understanding and reasoning over extended contexts.

Image-text retrieval is a fundamental task to bridge the semantic gap between natural language and vision. Recent works primarily focus on aligning textual meanings with visual appearance. However, they often overlook the semantic discrepancy caused by syntactic structure in natural language expressions and relationships among visual entities. This oversight would lead to sub-optimal alignment and degraded retrieval performance, since the underlying semantic dependencies and object interactions remain inadequately encoded in both textual and visual embeddings. In this paper, we propose a novel Visual-Linguistic Dependency Encoding (VL-DE) framework, which explicitly models the dependency information among textual words and interaction patterns between image regions, improving the discriminative power of cross-modal representations for more accurate image-text retrieval. Specifically, VL-DE enhances textual representations by considering syntactic relationships and dependency types, and visual representations by attending to its spatially neighboring regions. Cross-attention mechanism is then introduced to aggregate aligned region-word pairs into image-text similarities. Analysis on Winoground, a dataset specially designed to measure vision-linguistic compositional structure reasoning, shows that VL-DE outperforms existing methods, demonstrating its effectiveness at this task. Comprehensive experiments on two benchmarks, Flickr30K and MS-COCO, further validates the competitiveness of our approach.

In multidimensional dialogues, emotions serve not only as crucial mediators of emotional exchanges but also carry rich information. Therefore, accurately identifying the emotions of interlocutors and understanding the triggering factors of emotional changes are paramount. This study focuses on the tasks of multilingual dialogue emotion recognition and emotion reversal reasoning based on provocateurs, aiming to enhance the accuracy and depth of emotional understanding in dialogues. To achieve this goal, we propose a novel model, MBERT-TextRCNN-PL, designed to effectively capture emotional information of interlocutors. Additionally, we introduce XGBoost-EC (Emotion Capturer) to identify emotion provocateurs, thereby delving deeper into the causal relationships behind emotional changes. By comparing with state-of-the-art models, our approach demonstrates significant improvements in recognizing dialogue emotions and provocateurs, offering new insights and methodologies for multilingual dialogue emotion understanding and emotion reversal research.

Achieving empathy is a crucial step toward humanized dialogue systems. Current approaches for empathetic dialogue generation mainly perceive an emotional label to generate an empathetic response conditioned on it, which simply treat emotions independently, but ignore the intrinsic emotion correlation in dialogues, resulting in inaccurate emotion perception and unsuitable response generation. In this paper, we propose a novel emotion correlation enhanced empathetic dialogue generation framework, which comprehensively realizes emotion correlation learning, utilization, and supervising. Specifically, a multi-resolution emotion graph is devised to capture context-based emotion interactions from different resolutions, further modeling emotion correlation. Then we propose an emotion correlation enhanced decoder, with a novel correlation-aware aggregation and soft/hard strategy, respectively improving the emotion perception and response generation. Experimental results on the benchmark dataset demonstrate the superiority of our model in both empathetic perception and expression.

Extractive text summarisation aims to select salient sentences from a document to form a short yet informative summary. While learning-based methods have achieved promising results, they have several limitations, such as dependence on expensive training and lack of interpretability. Therefore, in this paper, we propose a novel non-learning-based method by for the first time formulating text summarisation as an Optimal Transport (OT) problem, namely Optimal Transport Extractive Summariser (OTExtSum). Optimal sentence extraction is conceptualised as obtaining an optimal summary that minimises the transportation cost to a given document regarding their semantic distributions. Such a cost is defined by the Wasserstein distance and used to measure the summary’s semantic coverage of the original document. Comprehensive experiments on four challenging and widely used datasets - MultiNews, PubMed, BillSum, and CNN/DM demonstrate that our proposed method outperforms the state-of-the-art non-learning-based methods and several recent learning-based methods in terms of the ROUGE metric.

Fact verification is an essential tool to mitigate the spread of false information online, which has gained a widespread attention recently. However, a fact verification in the question-answering dialogue is still underexplored. In this paper, we propose a neural network based approach called question-answering dialogue based fact verification with mixture of experts (QaDialMoE). It exploits questions and evidence effectively in the verification process and can significantly improve the performance of fact verification. Specifically, we exploit the mixture of experts to focus on various interactions among responses, questions and evidence. A manager with an attention guidance module is implemented to guide the training of experts and assign a reasonable attention score to each expert. A prompt module is developed to generate synthetic questions that make our approach more generalizable. Finally, we evaluate the QaDialMoE and conduct a comparative study on three benchmark datasets. The experimental results demonstrate that our QaDialMoE outperforms previous approaches by a large margin and achieves new state-of-the-art results on all benchmarks. This includes the accuracy improvements on the HEALTHVER as 84.26%, the FAVIQ A dev set as 78.7%, the FAVIQ R dev set as 86.1%, test set as 86.0%, and the COLLOQUIAL as 89.5%. To our best knowledge, this is the first work to investigate a question-answering dialogue based fact verification, and achieves new state-of-the-art results on various benchmark datasets.

Due to the increasing concerns for data privacy, source-free unsupervised domain adaptation attracts more and more research attention, where only a trained source model is assumed to be available, while the labeled source data remain private. To get promising adaptation results, we need to find effective ways to transfer knowledge learned in source domain and leverage useful domain specific information from target domain at the same time. This paper describes our winning contribution to SemEval 2021 Task 10: Source-Free Domain Adaptation for Semantic Processing. Our key idea is to leverage the model trained on source domain data to generate pseudo labels for target domain samples. Besides, we propose Negation-aware Pre-training (NAP) to incorporate negation knowledge into model. Our method win the 1st place with F1-score of 0.822 on the official blind test set of Negation Detection Track.

Faceted summarization provides briefings of a document from different perspectives. Readers can quickly comprehend the main points of a long document with the help of a structured outline. However, little research has been conducted on this subject, partially due to the lack of large-scale faceted summarization datasets. In this study, we present FacetSum, a faceted summarization benchmark built on Emerald journal articles, covering a diverse range of domains. Different from traditional document-summary pairs, FacetSum provides multiple summaries, each targeted at specific sections of a long document, including the purpose, method, findings, and value. Analyses and empirical results on our dataset reveal the importance of bringing structure into summaries. We believe FacetSum will spur further advances in summarization research and foster the development of NLP systems that can leverage the structured information in both long texts and summaries.

Pretrained Language Models (PLMs) have improved the performance of natural language understanding in recent years. Such models are pretrained on large corpora, which encode the general prior knowledge of natural languages but are agnostic to information characteristic of downstream tasks. This often results in overfitting when fine-tuned with low resource datasets where task-specific information is limited. In this paper, we integrate label information as a task-specific prior into the self-attention component of pretrained BERT models. Experiments on several benchmarks and real-word datasets suggest that the proposed approach can largely improve the performance of pretrained models when fine-tuning with small datasets.

Continuous efforts have been devoted to language understanding (LU) for conversational queries with the fast and wide-spread popularity of voice assistants. In this paper, we first study the LU problem in the spatial domain, which is a critical problem for providing location-based services by voice assistants but is without in-depth investigation in existing studies. Spatial domain queries have several unique properties making them be more challenging for language understanding than common conversational queries, including lexical-similar but diverse intents and highly ambiguous words. Thus, a special tailored LU framework for spatial domain queries is necessary. To the end, a dataset was extracted and annotated based on the real-life queries from a voice assistant service. We then proposed a new multi-task framework that jointly learns the intent detection and entity linking tasks on the with invented hierarchical intent detection method and triple-scoring mechanism for entity linking. A specially designed spatial GCN is also utilized to model spatial context information among entities. We have conducted extensive experimental evaluations with state-of-the-art entity linking and intent detection methods, which demonstrated that can outperform all baselines with a significant margin.

A reverse dictionary takes descriptions of words as input and outputs words semantically matching the input descriptions. Reverse dictionaries have great practical value such as solving the tip-of-the-tongue problem and helping new language learners. There have been some online reverse dictionary systems, but they support English reverse dictionary queries only and their performance is far from perfect. In this paper, we present a new open-source online reverse dictionary system named WantWords (https://wantwords.thunlp.org/). It not only significantly outperforms other reverse dictionary systems on English reverse dictionary performance, but also supports Chinese and English-Chinese as well as Chinese-English cross-lingual reverse dictionary queries for the first time. Moreover, it has user-friendly front-end design which can help users find the words they need quickly and easily. All the code and data are available at https://github.com/thunlp/WantWords.

Popular metrics used for evaluating image captioning systems, such as BLEU and CIDEr, provide a single score to gauge the system’s overall effectiveness. This score is often not informative enough to indicate what specific errors are made by a given system. In this study, we present a fine-grained evaluation method REO for automatically measuring the performance of image captioning systems. REO assesses the quality of captions from three perspectives: 1) Relevance to the ground truth, 2) Extraness of the content that is irrelevant to the ground truth, and 3) Omission of the elements in the images and human references. Experiments on three benchmark datasets demonstrate that our method achieves a higher consistency with human judgments and provides more intuitive evaluation results than alternative metrics.

This paper presents a new metric called TIGEr for the automatic evaluation of image captioning systems. Popular metrics, such as BLEU and CIDEr, are based solely on text matching between reference captions and machine-generated captions, potentially leading to biased evaluations because references may not fully cover the image content and natural language is inherently ambiguous. Building upon a machine-learned text-image grounding model, TIGEr allows to evaluate caption quality not only based on how well a caption represents image content, but also on how well machine-generated captions match human-generated captions. Our empirical tests show that TIGEr has a higher consistency with human judgments than alternative existing metrics. We also comprehensively assess the metric’s effectiveness in caption evaluation by measuring the correlation between human judgments and metric scores.

In this paper, we introduce our cross-lingual linked data lexica, called xLiD-Lexica, which are constructed by exploiting the multilingual Wikipedia and linked data resources from Linked Open Data (LOD). We provide the cross-lingual groundings of linked data resources from LOD as RDF data, which can be easily integrated into the LOD data sources. In addition, we build a SPARQL endpoint over our xLiD-Lexica to allow users to easily access them using SPARQL query language. Multilingual and cross-lingual information access can be facilitated by the availability of such lexica, e.g., allowing for an easy mapping of natural language expressions in different languages to linked data resources from LOD. Many tasks in natural language processing, such as natural language generation, cross-lingual entity linking, text annotation and question answering, can benefit from our xLiD-Lexica.

In recent years large repositories of structured knowledge (DBpedia, Freebase, YAGO) have become a valuable resource for language technologies, especially for the automatic aggregation of knowledge from textual data. One essential component of language technologies, which leverage such knowledge bases, is the linking of words or phrases in specific text documents with elements from the knowledge base (KB). We call this semantic annotation. In the same time, initiatives like Wikidata try to make those knowledge bases less language dependent in order to allow cross-lingual or language independent knowledge access. This poses a new challenge to semantic annotation tools which typically are language dependent and link documents in one language to a structured knowledge base grounded in the same language. Ultimately, the goal is to construct cross-lingual semantic annotation tools that can link words or phrases in one language to a structured knowledge database in any other language or to a language independent representation. To support this line of research we developed what we believe could serve as a gold standard Resource for Evaluating Cross-lingual Semantic Annotation (RECSA). We compiled a hand-annotated parallel corpus of 300 news articles in three languages with cross-lingual semantic groundings to the English Wikipedia and DBPedia. We hope that this new language resource, which is freely available, will help to establish a standard test set and methodology to comparatively evaluate cross-lingual semantic annotation technologies.

Notre travail concerne l’analyse automatique des énoncés d’opinion en chinois. En nous inspirant de la théorie linguistique de l’Appraisal, nous proposons une méthode fondée sur l’usage de lexiques et de règles locales pour déterminer les caractéristiques telles que la Force (intensité), le Focus (prototypicalité) et la polarité de tels énoncés. Nous présentons le modèle et sa mise en oeuvre sur un corpus journalistique. Si pour la détection d’énoncés d’opinion, la précision est bonne (94 %), le taux de rappel (67 %) pose cependant des questions sur l’enrichissement des ressources actuelles.