Mixture-of-Experts (MoE) has gained increasing popularity as a promising framework for scaling up large language models (LLMs). However, training MoE from scratch in a large-scale setting still suffers from data-hungry and instability problems. Motivated by this limit, we investigate building MoE models from existing dense large language models. Specifically, based on the well-known LLaMA-2 7B model, we obtain an MoE model by: (1) Expert Construction, which partitions the parameters of original Feed-Forward Networks (FFNs) into multiple experts; (2) Continual pre-training, which further trains the transformed MoE model and additional gate networks. In this paper, we comprehensively explore different methods for expert construction and various data sampling strategies for continual pre-training. After these stages, our LLaMA-MoE models could maintain language abilities and route the input tokens to specific experts with part of the parameters activated. Empirically, by training 200B tokens, LLaMA-MoE-3.5B models significantly outperform dense models that contain similar activation parameters.
Large Language Models (LLMs) have shown their impressive capabilities, while also raising concerns about the data contamination problems due to privacy issues and leakage of benchmark datasets in the pre-training phase. Therefore, it is vital to detect the contamination by checking whether an LLM has been pre-trained on the target texts. Recent studies focus on the generated texts and compute perplexities, which are superficial features and not reliable. In this study, we propose to utilize the probing technique for pre-training data detection by examining the model’s internal activations. Our method is simple and effective and leads to more trustworthy pre-training data detection. Additionally, we propose ArxivMIA, a new challenging benchmark comprising arxiv abstracts from Computer Science and Mathematics categories. Our experiments demonstrate that our method outperforms all baselines, and achieves state-of-the-art performance on both WikiMIA and ArxivMIA, with additional experiments confirming its efficacy.
Temporal reasoning is fundamental for large language models (LLMs) to comprehend the world. Current temporal reasoning datasets are limited to questions about single or isolated events, falling short in mirroring the realistic temporal characteristics involving concurrent nature and intricate temporal interconnections. In this paper, we introduce CoTempQA, a comprehensive co-temporal Question Answering (QA) benchmark containing four co-temporal scenarios (Equal, Overlap, During, Mix) with 4,748 samples for evaluating the co-temporal comprehension and reasoning abilities of LLMs. Our extensive experiments reveal a significant gap between the performance of current LLMs and human-level reasoning on CoTempQA tasks. Even when enhanced with Chain of Thought (CoT) methodologies, models consistently struggle with our task. In our preliminary exploration, we discovered that mathematical reasoning plays a significant role in handling co-temporal events and proposed a strategy to boost LLMs’ co-temporal reasoning from a mathematical perspective. We hope that our CoTempQA datasets will encourage further advancements in improving the co-temporal reasoning capabilities of LLMs.
Zero-shot dialogue state tracking (DST) transfers knowledge to unseen domains, reducing the cost of annotating new datasets. Previous zero-shot DST models mainly suffer from domain transferring and partial prediction problems. To address these challenges, we propose Mixture of Prefix Experts (MoPE) to establish connections between similar slots in different domains, which strengthens the model transfer performance in unseen domains. Empirical results demonstrate that MoPE-DST achieves the joint goal accuracy of 57.13% on MultiWOZ2.1 and 55.4.
Sharing knowledge between information extraction tasks has always been a challenge due to the diverse data formats and task variations. Meanwhile, this divergence leads to information waste and increases difficulties in building complex applications in real scenarios. Recent studies often formulate IE tasks as a triplet extraction problem. However, such a paradigm does not support multi-span and n-ary extraction, leading to weak versatility. To this end, we reorganize IE problems into unified multi-slot tuples and propose a universal framework for various IE tasks, namely Mirror. Specifically, we recast existing IE tasks as a multi-span cyclic graph extraction problem and devise a non-autoregressive graph decoding algorithm to extract all spans in a single step. It is worth noting that this graph structure is incredibly versatile, and it supports not only complex IE tasks, but also machine reading comprehension and classification tasks. We manually construct a corpus containing 57 datasets for model pretraining, and conduct experiments on 30 datasets across 8 downstream tasks. The experimental results demonstrate that our model has decent compatibility and outperforms or reaches competitive performance with SOTA systems under few-shot and zero-shot settings. The code, model weights, and pretraining corpus are available at https://github.com/Spico197/Mirror .
“多标签文本分类((Multi-Label Text Classification, MLTC)旨在从预定义的候选标签集合中选择一个或多个文本对应的类别,是自然语言处理C)旨在从预定义的候选标签集合中选择一个或多个文本对应的类别,是自然语言处理(Natural Language Processing,NLP)的一项基本任务。前人工作大多基于规范且全面的标注数据集,而这些规范数据集需要严格的质量控制,一般很难获取。在真实的标注过程中,难免会丢失掉一些相关标签,进而导致不完全标注问题。为此本文提出了一种基于局部标注的自监督框架(Partial Self-Training,PST),该框架利用教师模型自动地给大规模无标注数据打伪标签,同时给不完全标注数据补充缺失标签,最后再利用这些数据反向更新教师模型。在合成数据集和真实数据集上的实验表明,本文提出的PST框架兼容现有的各类多标签文本分类模型,并且可以缓解不完全标注数据对模型的影响。”
Supervised models for Relation Extraction (RE) typically require human-annotated training data. Due to the limited size, the human-annotated data is usually incapable of covering diverse relation expressions, which could limit the performance of RE. To increase the coverage of relation expressions, we may enlarge the labeled data by hiring annotators or applying Distant Supervision (DS). However, the human-annotated data is costly and non-scalable while the distantly supervised data contains many noises. In this paper, we propose an alternative approach to improve RE systems via enriching diverse expressions by relational paraphrase sentences. Based on an existing labeled data, we first automatically build a task-specific paraphrase data. Then, we propose a novel model to learn the information of diverse relation expressions. In our model, we try to capture this information on the paraphrases via a joint learning framework. Finally, we conduct experiments on a widely used dataset and the experimental results show that our approach is effective to improve the performance on relation extraction, even compared with a strong baseline.
In recent years, distantly-supervised relation extraction has achieved a certain success by using deep neural networks. Distant Supervision (DS) can automatically generate large-scale annotated data by aligning entity pairs from Knowledge Bases (KB) to sentences. However, these DS-generated datasets inevitably have wrong labels that result in incorrect evaluation scores during testing, which may mislead the researchers. To solve this problem, we build a new dataset NYTH, where we use the DS-generated data as training data and hire annotators to label test data. Compared with the previous datasets, NYT-H has a much larger test set and then we can perform more accurate and consistent evaluation. Finally, we present the experimental results of several widely used systems on NYT-H. The experimental results show that the ranking lists of the comparison systems on the DS-labelled test data and human-annotated test data are different. This indicates that our human-annotated data is necessary for evaluation of distantly-supervised relation extraction.