The rapid advancement of large language models (LLMs) has sparked interest in data synthesis techniques, aiming to generate diverse and high-quality synthetic datasets. However, these synthetic datasets often suffer from a lack of diversity and added noise. In this paper, we present TarGEN, a multi-step prompting strategy for generating high-quality synthetic datasets utilizing a LLM. An advantage of TarGEN is its seedless nature; it does not require specific task instances, broadening its applicability beyond task replication. We augment TarGEN with a method known as self-correction empowering LLMs to rectify inaccurately labeled instances during dataset creation, ensuring reliable labels. To assess our technique's effectiveness, we emulate 8 tasks from the SuperGLUE benchmark and finetune various language models, including encoder-only, encoder-decoder, and decoder-only models on both synthetic and original training sets. Evaluation on the original test set reveals that models trained on datasets generated by TarGEN perform approximately 1-2% points better than those trained on original datasets (82.84% via syn. vs. 81.12% on og. using Flan-T5). When incorporating instruction tuning, the performance increases to 84.54% on synthetic data vs. 81.49% on original data by Flan-T5. A comprehensive analysis of the synthetic dataset compared to the original dataset reveals that the synthetic dataset demonstrates similar or higher levels of dataset complexity and diversity. Furthermore, the synthetic dataset displays a bias level that aligns closely with the original dataset. Finally, when pre-finetuned on our synthetic SuperGLUE dataset, T5-3B yields impressive results on the OpenLLM leaderboard, surpassing the model trained on the Self-Instruct dataset by 4.14% points. We hope that TarGEN can be helpful for quality data generation and reducing the human efforts to create complex benchmarks.
CraftRTL: High-quality Synthetic Data Generation for Verilog Code Models with Correct-by-Construction Non-Textual Representations and Targeted Code Repair
From General to Targeted Rewards: Surpassing GPT-4 in Open-Ended Long-Context Generation
The Chinese University of Hong Kong（香港中文大学）；National University of Singapore（新加坡国立大学）；Nanyang Technological University（南洋理工大学）；Macquarie University（麦考瑞大学）
TARGA: Targeted Synthetic Data Generation for Practical Reasoning over Structured Data
State Key Laboratory for Novel Software Technology, Nanjing University（南京大学计算机软件新技术国家重点实验室）；University of California, Santa Barbara（加州大学圣巴巴拉分校）；University of Oxford（牛津大学）
Existing language models excel at writing from scratch, but many real-world scenarios require rewriting an existing document to fit a set of constraints. Although sentence-level rewriting has been fairly well-studied, little work has addressed the challenge of rewriting an entire document coherently. In this work, we introduce the task of document-level targeted content transfer and address it in the recipe domain, with a recipe as the document and a dietary restriction (such as vegan or dairy-free) as the targeted constraint. We propose a novel model for this task based on the generative pre-…
EtCon: Edit-then-Consolidate for Reliable Knowledge Editing
Wuhan University（武汉大学）；Shanghai Innovation Institute（上海创新研究院）；Fudan University（复旦大学）；Shanghai Jiao Tong University（上海交通大学）；University of Science and Technology of China（中国科学技术大学）
Previous work has shown that Large Language Models are susceptible to so-called data extraction attacks. This allows an attacker to extract a sample that was contained in the training data, which has massive privacy implications. The construction of data extraction attacks is challenging, current attacks are quite inefficient, and there exists a significant gap in the extraction capabilities of untargeted attacks and memorization. Thus, targeted attacks are proposed, which identify if a given sample from the training data, is extractable from a model. In this work, we apply a targeted data ext…
Can LLMs Generate Reliable Test Case Generators? A Study on Competition-Level Programming Problems
Shanghai Qi Zhi Institute（上海启智研究院）；ShanghaiTech University（上海科技大学）；Wuhan University（武汉大学）；Fuzhou University（福州大学）；Institute of Information Engineering, Chinese Academy of Sciences（中国科学院信息工程研究所）；Tsinghua University（清华大学）；Huazhong University of Science and Technology（华中科技大学）；Nanjing University（南京大学）；Beijing University of Posts and Telecommunications（北京邮电大学）；Peking University（北京大学）
Intelligent and adaptive online education systems aim to make high-quality education available for a diverse range of students. However, existing systems usually depend on a pool of hand-made questions, limiting how fine-grained and open-ended they can be in adapting to individual students. We explore targeted question generation as a controllable sequence generation task. We first show how to fine-tune pre-trained language models for deep knowledge tracing (LM-KT). This model accurately predicts the probability of a student answering a question correctly, and generalizes to questions not seen…
A major obstacle to the wide-spread adoption of neural retrieval models is that they require large supervised training sets to surpass traditional term-based techniques, which are constructed from raw corpora. In this paper, we propose an approach to zero-shot learning for passage retrieval that uses synthetic question generation to close this gap. The question generation system is trained on general domain data, but is applied to documents in the targeted domain. This allows us to create arbitrarily large, yet noisy, question-passage relevance pairs that are domain specific. Furthermore, when…
With an increasing number of malicious attacks, the number of people and organizations falling prey to social engineering attacks is proliferating. Despite considerable research in mitigation systems, attackers continually improve their modus operandi by using sophisticated machine learning, natural language processing techniques with an intent to launch successful targeted attacks aimed at deceiving detection mechanisms as well as the victims. We propose a system for advanced email masquerading attacks using Natural Language Generation (NLG) techniques. Using legitimate as well as an influx o…
Stories generated with neural language models have shown promise in grammatical and stylistic consistency. However, the generated stories are still lacking in common sense reasoning, e.g., they often contain sentences deprived of world knowledge. We propose a simple multi-task learning scheme to achieve quantitatively better common sense reasoning in language models by leveraging auxiliary training signals from datasets designed to provide common sense grounding. When combined with our two-stage fine-tuning pipeline, our method achieves improved common sense reasoning and state-of-the-art perp…
Westlake University（西湖大学）；Westlake Institute for Advanced Study（西湖高等研究院）；Nanjing University of Science and Technology（南京理工大学）；National University of Singapore（新加坡国立大学）
Since previous studies on open-domain targeted sentiment analysis are limited in dataset domain variety and sentence level, we propose a novel dataset consisting of 6,013 human-labeled data to extend the data domains in topics of interest and document level. Furthermore, we offer a nested target annotation schema to extract the complete sentiment information in documents, boosting the practicality and effectiveness of open-domain targeted sentiment analysis. Moreover, we leverage the pre-trained model BART in a sequence-to-sequence generation method for the task. Benchmark results show that th…
Advanced machine learning and natural language techniques enable attackers to launch sophisticated and targeted social engineering-based attacks. To counter the active attacker issue, researchers have since resorted to proactive methods of detection. Email masquerading using targeted emails to fool the victim is an advanced attack method. However automatic text generation requires controlling the context and coherency of the generated content, which has been identified as an increasingly difficult problem. The method used leverages a hierarchical deep neural model which uses a learned represen…
Good quality explanations of artificial intelligence (XAI) reasoning must be written (and evaluated) for an explanatory purpose, targeted towards their readers, have a good narrative and causal structure, and highlight where uncertainty and data quality affect the AI output. I discuss these challenges from a Natural Language Generation (NLG) perspective, and highlight four specific NLG for XAI research challenges.
University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；Zhejiang University（浙江大学）；Tencent（腾讯）；Microsoft Dynamics 365 AI Research（微软Dynamics 365人工智能研究院）
Adversarial attacks against natural language processing systems, which perform seemingly innocuous modifications to inputs, can induce arbitrary mistakes to the target models. Though raised great concerns, such adversarial attacks can be leveraged to estimate the robustness of NLP models. Compared with the adversarial example generation in continuous data domain (e.g., image), generating adversarial text that preserves the original meaning is challenging since the text space is discrete and non-differentiable. To handle these challenges, we propose a target-controllable adversarial attack fram…
Due to its expressivity, natural language is paramount for explicit and implicit affective state communication among humans. The same linguistic inquiry (e.g., How are you?) might induce responses with different affects depending on the affective state of the conversational partner(s) and the context of the conversation. Yet, most dialog systems do not consider affect as constitutive aspect of response generation. In this paper, we introduce AffectON, an approach for generating affective responses during inference. For generating language in a targeted affect, our approach leverages a probabil…
LLM-Bootstrapped Targeted Finding Guidance for Factual MLLM-based Medical Report Generation
Zhejiang University（浙江大学）；The Second Affiliated Hospital Zhejiang University School of Medicine（浙江大学医学院附属第二医院）；Hangzhou Pu Jian Medical Technology Co., Ltd.（杭州普健医疗科技有限公司）
Motivation: The development of novel compounds targeting proteins of interest is one of the most important tasks in the pharmaceutical industry. Deep generative models have been applied to targeted molecular design and have shown promising results. Recently, target-specific molecule generation has been viewed as a translation between the protein language and the chemical language. However, such a model is limited by the availability of interacting protein-ligand pairs. On the other hand, large amounts of unlabeled protein sequences and chemical compounds are available and have been used to tra…
Microsoft Research（微软研究院）；Alexa AI, Amazon（亚马逊 Alexa AI）
Generating responses in a targeted style is a useful yet challenging task, especially in the absence of parallel data. With limited data, existing methods tend to generate responses that are either less stylized or less context-relevant. We propose StyleFusion, which bridges conversation modeling and non-parallel style transfer by sharing a structured latent space. This structure allows the system to generate stylized relevant responses by sampling in the neighborhood of the conversation model prediction, and continuously control the style level. We demonstrate this method using dialogues from…
Charles University（查理大学）；Institute of Formal and Applied Linguistics（形式与应用语言学研究所）
We present the first dataset targeted at end-to-end NLG in Czech in the restaurant domain, along with several strong baseline models using the sequence-to-sequence approach. While non-English NLG is under-explored in general, Czech, as a morphologically rich language, makes the task even harder: Since Czech requires inflecting named entities, delexicalization or copy mechanisms do not work out-of-the-box and lexicalizing the generated outputs is non-trivial. In our experiments, we present two different approaches to this this problem: (1) using a neural language model to select the correct inf…
University of North Carolina, Chapel Hill（北卡罗来纳大学教堂山分校）
In this paper, we introduce a new dataset consisting of 360,001 focused natural language descriptions for 10,738 images. This dataset, the Visual Madlibs dataset, is collected using automatically produced fill-in-the-blank templates designed to gather targeted descriptions about: people and objects, their appearances, activities, and interactions, as well as inferences about the general scene or its broader context. We provide several analyses of the Visual Madlibs dataset and demonstrate its applicability to two new description generation tasks: focused description generation, and multiple-ch…
On Evaluating Adversarial Robustness of Large Vision-Language Models
Tsinghua University（清华大学）；Renmin University of China（中国人民大学）；Sea AI Lab, Singapore（新加坡Sea AI实验室）；Singapore University of Technology and Design（新加坡科技设计大学）
