Self-Alignment for Factuality: Mitigating Hallucinations in LLMs via Self-Evaluation

Despite showing increasingly human-like abilities, large language models (LLMs) often struggle with factual inaccuracies, i.e. "hallucinations", even when they hold relevant knowledge. To address these hallucinations, current approaches typically necessitate high-quality human factuality annotations. In this work, we explore Self-Alignment for Factuality, where we leverage the self-evaluation capability of an LLM to provide training signals that steer the model towards factuality. Specifically, we incorporate Self-Eval, a self-evaluation component, to prompt an LLM to validate the factuality of its own generated responses solely based on its internal knowledge. Additionally, we design Self-Knowledge Tuning (SK-Tuning) to augment the LLM's self-evaluation ability by improving the model's confidence estimation and calibration. We then utilize these self-annotated responses to fine-tune the model via Direct Preference Optimization algorithm. We show that the proposed self-alignment approach substantially enhances factual accuracy over Llama family models across three key knowledge-intensive tasks on TruthfulQA and BioGEN.

A Survey of the Evolution of Language Model-Based Dialogue Systems

Dialogue systems, including task-oriented_dialogue_system (TOD) and open-domain_dialogue_system (ODD), have undergone significant transformations, with language_models (LM) playing a central role. This survey delves into the historical trajectory of dialogue systems, elucidating their intricate relationship with advancements in language models by categorizing this evolution into four distinct stages, each marked by pivotal LM breakthroughs: 1) Early_Stage: characterized by statistical LMs, resulting in rule-based or machine-learning-driven dialogue_systems; 2) Independent development of TOD and ODD based on neural_language_models (NLM; e.g., LSTM and GRU), since NLMs lack intrinsic knowledge in their parameters; 3) fusion between different types of dialogue systems with the advert of pre-trained_language_models (PLMs), starting from the fusion between four_sub-tasks_within_TOD, and then TOD_with_ODD; and 4) current LLM-based_dialogue_system, wherein LLMs can be used to conduct TOD and ODD seamlessly. Thus, our survey provides a chronological perspective aligned with LM breakthroughs, offering a comprehensive review of state-of-the-art research outcomes. What's more, we focus on emerging topics and discuss open challenges, providing valuable insights into future directions for LLM-based_dialogue_systems. Through this exploration, we pave the way for a deeper_comprehension of the evolution, guiding future developments in LM-based dialogue_systems.

Rethinking Machine Ethics -- Can LLMs Perform Moral Reasoning through the Lens of Moral Theories?

Making moral judgments is an essential step toward developing ethical AI systems. Prevalent approaches are mostly implemented in a bottom-up manner, which uses a large set of annotated data to train models based on crowd-sourced opinions about morality. These approaches have been criticized for potentially overgeneralizing a limited group of annotators' moral stances and lacking explainability. In contrast, top-down approaches make moral judgments grounded in a set of principles. However, it remains conceptual due to the incapability of previous language models and the unsolved debate among moral principles. In this study, we propose a flexible framework to steer Large Language Models (LLMs) to perform moral reasoning with well-established moral theories from interdisciplinary research. The theory-guided top-down framework can incorporate various moral theories. Our experiments demonstrate the effectiveness of the proposed framework on datasets derived from moral theories. Furthermore, we show the alignment between different moral theories and existing morality datasets. Our analysis exhibits the potentials and flaws in existing resources (models and datasets) in developing explainable moral judgment-making systems.

SGP-TOD: Building Task Bots Effortlessly via Schema-Guided LLM Prompting

Building end-to-end task bots and maintaining their integration with new functionalities using minimal human efforts is a long-standing challenge in dialog research. Recently large language models (LLMs) have demonstrated exceptional proficiency in conversational engagement and adherence to instructions across various downstream tasks. In this work, we introduce SGP-TOD, Schema-Guided Prompting for building Task-Oriented Dialog systems effortlessly based on LLMs. Utilizing the symbolic knowledge -- task schema, we instruct fixed LLMs to generate appropriate responses on novel tasks, circumventing the need for training data. Specifically, SGP-TOD comprises three components: a LLM for engaging with users, a DST Prompter to aid the LLM with dialog state tracking, which is then used to retrieve database items, and a Policy Prompter to elicit proper responses adhering to the provided dialog policy. Experimental results on Multiwoz, RADDLE and STAR datasets show that our training-free strategy SGP-TOD, without any task-specific data, yields state-of-the-art (SOTA) zero-shot performance, greatly surpasses the few-shot approaches. In a domain-extension setting, SGP-TOD aptly adapts to new functionalities by merely adding supplementary schema rules. We make our code and data publicly available.

PanGu-Bot: Efficient Generative Dialogue Pre-training from Pre-trained Language Model

In this paper, we introduce PanGu-Bot, a Chinese pre-trained open-domain dialogue generation model based on a large pre-trained language model (PLM) PANGU-alpha (Zeng et al.,2021). Different from other pre-trained dialogue models trained over a massive amount of dialogue data from scratch, we aim to build a powerful dialogue model with relatively fewer data and computation costs by inheriting valuable language capabilities and knowledge from PLMs. To this end, we train PanGu-Bot from the large PLM PANGU-alpha, which has been proven well-performed on a variety of Chinese natural language tasks. We investigate different aspects of responses generated by PanGu-Bot, including response quality, knowledge, and safety. We show that PanGu-Bot outperforms state-of-the-art Chinese dialogue systems (CDIALGPT (Wang et al., 2020), EVA (Zhou et al., 2021)) w.r.t. the above three aspects. We also demonstrate that PanGu-Bot can be easily deployed to generate emotional responses without further training. Throughout our empirical analysis, we also point out that the PanGu-Bot response quality, knowledge correctness, and safety are still far from perfect, and further explorations are indispensable to building reliable and smart dialogue systems. Our model and code will be available at https://github.com/huawei-noah/Pretrained-Language-Model/tree/master/PanGu-Bot soon.

Towards Identifying Social Bias in Dialog Systems: Frame, Datasets, and Benchmarks

The research of open-domain dialog systems has been greatly prospered by neural models trained on large-scale corpora, however, such corpora often introduce various safety problems (e.g., offensive languages, biases, and toxic behaviors) that significantly hinder the deployment of dialog systems in practice. Among all these unsafe issues, addressing social bias is more complex as its negative impact on marginalized populations is usually expressed implicitly, thus requiring normative reasoning and rigorous analysis. In this paper, we focus our investigation on social bias detection of dialog safety problems. We first propose a novel Dial-Bias Frame for analyzing the social bias in conversations pragmatically, which considers more comprehensive bias-related analyses rather than simple dichotomy annotations. Based on the proposed framework, we further introduce CDail-Bias Dataset that, to our knowledge, is the first well-annotated Chinese social bias dialog dataset. In addition, we establish several dialog bias detection benchmarks at different label granularities and input types (utterance-level and context-level). We show that the proposed in-depth analyses together with these benchmarks in our Dial-Bias Frame are necessary and essential to bias detection tasks and can benefit building safe dialog systems in practice.

Convex Polytope Modelling for Unsupervised Derivation of Semantic Structure for Data-efficient Natural Language Understanding

Popular approaches for Natural Language Understanding (NLU) usually rely on a huge amount of annotated data or handcrafted rules, which is laborious and not adaptive to domain extension. We recently proposed a Convex-Polytopic-Model-based framework that shows great potential in automatically extracting semantic patterns by exploiting the raw dialog corpus. The extracted semantic patterns can be used to generate semantic frames, which is essential in assisting NLU tasks. This paper further studies the CPM model in depth and visualizes its high interpretability and transparency at various levels. We show that this framework can exploit semantic-frame-related features in the corpus, reveal the underlying semantic structure of the utterances, and boost the performance of the state-of-the-art NLU model with minimal supervision. We conduct our experiments on the ATIS (Air Travel Information System) corpus.

COLD: A Benchmark for Chinese Offensive Language Detection

Offensive language detection and prevention becomes increasing critical for maintaining a healthy social platform and the safe deployment of language models. Despite plentiful researches on toxic and offensive language problem in NLP, existing studies mainly focus on English, while few researches involve Chinese due to the limitation of resources. To facilitate Chinese offensive language detection and model evaluation, we collect COLDataset, a Chinese offensive language dataset containing 37k annotated sentences. With this high-quality dataset, we provide a strong baseline classifier, COLDetector, with 81% accuracy for offensive language detection. Furthermore, we also utilize the proposed \textsc{COLDetector} to study output offensiveness of popular Chinese language models (CDialGPT and CPM). We find that (1) CPM tends to generate more offensive output than CDialGPT, and (2) certain type of prompts, like anti-bias sentences, can trigger offensive outputs more easily.Altogether, our resources and analyses are intended to help detoxify the Chinese online communities and evaluate the safety performance of generative language models. Disclaimer: The paper contains example data that may be considered profane, vulgar, or offensive.

Unstructured Knowledge Access in Task-oriented Dialog Modeling using Language Inference, Knowledge Retrieval and Knowledge-Integrative Response Generation

Dialog systems enriched with external knowledge can handle user queries that are outside the scope of the supporting databases/APIs. In this paper, we follow the baseline provided in DSTC9 Track 1 and propose three subsystems, KDEAK, KnowleDgEFactor, and Ens-GPT, which form the pipeline for a task-oriented dialog system capable of accessing unstructured knowledge. Specifically, KDEAK performs knowledge-seeking turn detection by formulating the problem as natural language inference using knowledge from dialogs, databases and FAQs. KnowleDgEFactor accomplishes the knowledge selection task by formulating a factorized knowledge/document retrieval problem with three modules performing domain, entity and knowledge level analyses. Ens-GPT generates a response by first processing multiple knowledge snippets, followed by an ensemble algorithm that decides if the response should be solely derived from a GPT2-XL model, or regenerated in combination with the top-ranking knowledge snippet. Experimental results demonstrate that the proposed pipeline system outperforms the baseline and generates high-quality responses, achieving at least 58.77% improvement on BLEU-4 score.

The JDDC Corpus: A Large-Scale Multi-Turn Chinese Dialogue Dataset for E-commerce Customer Service

Human conversations in real scenarios are complicated and building a human-like dialogue agent is an extremely challenging task. With the rapid development of deep learning techniques, data-driven models become more and more prevalent which need a huge amount of real conversation data. In this paper, we construct a large-scale real scenario Chinese E-commerce conversation corpus, JDDC, with more than 1 million multi-turn dialogues, 20 million utterances, and 150 million words. The dataset reflects several characteristics of human-human conversations, e.g., goal-driven, and long-term dependency among the context. It also covers various dialogue types including task-oriented, chitchat and question-answering. Extra intent information and three well-annotated challenge sets are also provided. Then, we evaluate several retrieval-based and generative models to provide basic benchmark performance on JDDC corpus. And we hope JDDC can serve as an effective testbed and benefit the development of fundamental research in dialogue task.