This tutorial offers a comprehensive guide on using multi-armed bandit (MAB) algorithms to improve Large Language Models (LLMs). As Natural Language Processing (NLP) tasks grow, efficient and adaptive language generation systems are increasingly needed. MAB algorithms, which balance exploration and exploitation under uncertainty, are promising for enhancing LLMs. The tutorial covers foundational MAB concepts, including the exploration-exploitation trade-off and strategies like epsilon-greedy, UCB (Upper Confidence Bound), and Thompson Sampling. It then explores integrating MAB with LLMs, focusing on designing architectures that treat text generation options as arms in a bandit problem. Practical aspects like reward design, exploration policies, and scalability are discussed. Real-world case studies demonstrate the benefits of MAB-augmented LLMs in content recommendation, dialogue generation, and personalized content creation, showing how these techniques improve relevance, diversity, and user engagement.
论文检索
输入标题、作者或关键词，从 1,444 篇学术成果中精准定位
Lookahead: An Inference Acceleration Framework for Large Language Model with Lossless Generation Accuracy
PDF ↗As Large Language Models (LLMs) have made significant advancements across various tasks, such as question answering, translation, text summarization, and dialogue systems, the need for accuracy in information becomes crucial, especially for serious financial products serving billions of users like Alipay. However, for a real-world product serving millions of users, the inference speed of LLMs becomes a critical factor compared to a mere experimental model. Hence, this paper presents a generic framework for accelerating the inference process, resulting in a substantial increase in speed and cost reduction for our LLM-based scenarios, with lossless generation accuracy. In the traditional inference process, each token is generated sequentially by the LLM, leading to a time consumption proportional to the number of generated tokens. To enhance this process, our framework, named lookahead, introduces a multi-branch strategy. Instead of generating a single token at a time, we propose a Trie-based retrieval and verification mechanism to be able to accept several tokens at a forward step. Our strategy offers two distinct advantages: (1) it guarantees absolute correctness of the output, avoiding any approximation algorithms, and (2) the worst-case performance of our approach could be comparable with the performance of the conventional process. We conduct extensive experiments to demonstrate the significant improvements achieved by applying our inference acceleration framework. Our framework has been widely deployed in Alipay since April 2023, and obtained remarkable 2.66x to 6.26x speedup. Our code is available at https://github.com/alipay/PainlessInferenceAcceleration.
GRILLBot In Practice: Lessons and Tradeoffs Deploying Large Language Models for Adaptable Conversational Task Assistants
PDF ↗We tackle the challenge of building real-world multimodal assistants for complex real-world tasks. We describe the practicalities and challenges of developing and deploying GRILLBot, a leading (first and second prize winning in 2022 and 2023) system deployed in the Alexa Prize TaskBot Challenge. Building on our Open Assistant Toolkit (OAT) framework, we propose a hybrid architecture that leverages Large Language Models (LLMs) and specialised models tuned for specific subtasks requiring very low latency. OAT allows us to define when, how and which LLMs should be used in a structured and deployable manner. For knowledge-grounded question answering and live task adaptations, we show that LLM reasoning abilities over task context and world knowledge outweigh latency concerns. For dialogue state management, we implement a code generation approach and show that specialised smaller models have 84% effectiveness with 100x lower latency. Overall, we provide insights and discuss tradeoffs for deploying both traditional models and LLMs to users in complex real-world multimodal environments in the Alexa TaskBot challenge. These experiences will continue to evolve as LLMs become more capable and efficient -- fundamentally reshaping OAT and future assistant architectures.
Marrying Dialogue Systems with Data Visualization: Interactive Data Visualization Generation from Natural Language Conversations
PDF ↗Data visualization (DV) has become the prevailing tool in the market due to its effectiveness into illustrating insights in vast amounts of data. To lower the barrier of using DVs, automatic DV tasks, such as natural language question (NLQ) to visualization translation (formally called text-to-vis), have been investigated in the research community. However, text-to-vis assumes the NLQ to be well-organized and expressed in a single sentence. However, in real-world settings, complex DV is needed through consecutive exchanges between the DV system and the users. In this paper, we propose a new task named CoVis, short for Conversational text-to-Visualization, aiming at constructing DVs through a series of interactions between users and the system. Since it is the task which has not been studied in the literature, we first build a benchmark dataset named Dial-NVBench, including dialogue sessions with a sequence of queries from a user and responses from the system. The ultimate goal of each dialogue session is to create a suitable DV. However, this process can contain diverse dialogue queries, such as seeking information about the dataset, manipulating parts of the data, and visualizing the data. Then, we propose a multi-modal neural network named MMCoVisNet to answer these DV-related queries. In particular, MMCoVisNet first fully understands the dialogue context and determines the corresponding responses. Then, it uses adaptive decoders to provide the appropriate replies: (i) a straightforward text decoder is used to produce general responses, (ii) an SQL-form decoder is applied to synthesize data querying responses, and (iii) a DV-form decoder tries to construct the appropriate DVs. We comparatively evaluate MMCoVisNet with other baselines over our proposed benchmark dataset. Experimental results validate that MMCoVisNet performs better than existing baselines and achieves a state-of-the-art performance.
Many billion-scale large language models (LLMs) have been released for resource-constraint mobile devices to provide local LLM inference service when cloud-based powerful LLMs are not available. However, the capabilities of current on-device LLMs still lag behind those of cloud-based LLMs, and how to effectively and efficiently enhance on-device LLM inference becomes a practical requirement. We thus propose to collect the user's historical interactions with the cloud-based LLM and build an external datastore on the mobile device for enhancement using nearest neighbors search. Nevertheless, the full datastore improves the quality of token generation at the unacceptable expense of much slower generation speed. To balance performance and efficiency, we propose to select an optimal subset of the full datastore within the given size limit, the optimization objective of which is proven to be submodular. We further design an offline algorithm, which selects the subset after the construction of the full datastore, as well as an online algorithm, which performs selection over the stream and can be flexibly scheduled. We theoretically analyze the performance guarantee and the time complexity of the offline and the online designs to demonstrate effectiveness and scalability. We finally take three ChatGPT related dialogue datasets and four different on-device LLMs for evaluation. Evaluation results show that the proposed designs significantly enhance LLM performance in terms of perplexity while maintaining fast token generation speed. Practical overhead testing on the smartphone reveal the efficiency of on-device datastore subset selection from memory usage and computation overhead.
In this paper, we introduce the novel technological framework for the development of emotional dialogue systems. Inspired by the "Inside Out" film, we propose to use multiple emotional agents based on Large Language Models (LLMs) to prepare answers to a user query. Their answers are aggregated into a single response, taking into account the current emotional state of a user. The latter is estimated by video-based facial expression recognition (FER). We introduce several publicly available lightweight neural networks that show near state-of-the-art results on the AffectNet dataset. Qualitative examples using either GPT-3.5 or LLama2 and Mistral demonstrate that the proposed approach leads to more emotional responses in LLMs.
In contemporary society, the integration of artificial intelligence (AI) systems into various aspects of daily life raises significant ethical concerns. One critical aspect is to ensure that AI systems align with the moral values of the endusers. To that end, we introduce the Contextual Moral Value Alignment System, ComVas. Unlike traditional AI systems which have moral values predefined, ComVas empowers users to dynamically select and customize the desired moral values thereby guiding the system’s decision-making process. Through a user-friendly interface, individuals can specify their preferred morals, allowing the system to steer the model’s responses and actions accordingly. ComVas utilizes advanced natural language processing techniques to engage with the users in a meaningful dialogue, understanding their preferences, and reasoning about moral dilemmas in diverse contexts. This demo article showcases the functionality of ComVas, illustrating its potential to foster ethical decision-making in AI systems while respecting individual autonomy and promoting user-centric design principles.
ProMoAI is a novel tool that leverages Large Language Models (LLMs) to automatically generate process models from textual descriptions, incorporating advanced prompt engineering, error handling, and code generation techniques. Beyond automating the generation of complex process models, ProMoAI also supports process model optimization. Users can interact with the tool by providing feedback on the generated model, which is then used for refining the process model. ProMoAI utilizes the capabilities LLMs to offer a novel, AI-driven approach to process modeling, significantly reducing the barrier to entry for users without deep technical knowledge in process modeling.
Building an intelligent and efficient medical assistant is still a challenging AI problem. The major limitation comes from the data modality scarceness, which reduces comprehensive patient perception. This demo paper presents GigaPevt, the first multimodal medical assistant that combines the dialog capabilities of large language models with specialized medical models. Such an approach shows immediate advantages in dialog quality and metric performance, with a 1.18% accuracy improvement in the question-answering task.
Social learning plays an important role in the development of human intelligence. As children, we imitate our parent's speech patterns until we are able to produce sounds; we learn from them praising us and scolding us, and as adults, we learn by working with others. In this work, we survey the degree to which this developmental paradigm -- social learning -- has been mirrored in machine learning. In particular, since learning socially requires interacting with others, we are interested in how embodied agents can and have utilised these techniques. This is especially in light of the degree to which recent advances in natural language processing (NLP) enable us to perform new forms of social learning. We look at how behaviour cloning and next-token prediction mirror human imitation, how learning from human feedback mirrors human education, and how we can go further to enable fully communicative agents that learn from each other. We find that while individual social learning techniques have been used successfully, there has been little unifying work showing how to bring them together into socially embodied agents.
In safety-critical applications where a human is in the loop, providing timely contextual assistance can reduce the severity of emergencies. While the context can typically be inferred passively, engaging the human in an active conversation with the assistance system makes this context richer and more sound. For this, we explore a FOND-planning-powered goal-directed dialogue system with Natural Language Understanding (NLU) capabilities. We use an Ultralight (UL) aviation domain as an example application for test and validation by inferring the current context in situations requiring emergency landings using the goal-directed dialogue system. The inferred context is then used for real-time modelling of the problem instance, necessary for generating strategic plans to guide the human out of the emergency situations. To overcome data scarcity, we augment the data collected from human pilots using generative text models to train the NLU capabilities of the dialogue agent. We benchmark against generative chatbots and demonstrate that our goal-directed dialogue system significantly outperforms them in context inference.
Human-Agent Cooperation in Games under Incomplete Information through Natural Language Communication
PDF ↗Developing autonomous agents that can strategize and cooperate with humans under information asymmetry is challenging without effective communication in natural language. We introduce a shared-control game, where two players collectively control a token in alternating turns to achieve a common objective under incomplete information. We formulate a policy synthesis problem for an autonomous agent in this game with a human as the other player. To solve this problem, we propose a communication-based approach comprising a language module and a planning module. The language module translates natural language messages into and from a finite set of flags, a compact representation defined to capture player intents. The planning module leverages these flags to compute a policy using an asymmetric information-set Monte Carlo tree search with flag exchange algorithm we present. We evaluate the effectiveness of this approach in a testbed based on Gnomes at Night, a search-and-find maze board game. Results of human subject experiments show that communication narrows the information gap between players and enhances human-agent cooperation efficiency with fewer turns.
Decoupling Breaks Data Barriers: A Decoupled Pre-training Framework for Multi-intent Spoken Language Understanding
PDF ↗Multi-intent Spoken Language Understanding (Multi-intent SLU) can extract multiple intents in a single utterance, gaining increasing attention. Nevertheless, current multi-intent SLU approaches still heavily rely on large amounts of annotated multi-intent SLU data, which makes it hard to be satisfied in real-world scenarios without sufficient data. Motivated by this, we introduce a novel decoupled pre-training framework (DPF) to address the data-scarcity problem, achieving to leverage of abundant multi-intent-free SLU data to enhance multi-intent SLU. Specifically, DPF first decouples the multi-intent SLU task into two abilities: (1) task-agnostic ability to locate the task-agnostic slot entity span and (2) task-specific ability to predict the task-specific slot and intent labels simultaneously. The key insight of DPF is that such decomposition allows us to design a two-stage decoupled pre-training procedure to enhance both task-agnostic ability and task-specific ability with abundant multi-intent-free SLU data (i.e., NER and single-intent SLU data), respectively. Experimental results on two standard benchmarks (e.g., MixATIS and MixSNIPS) demonstrate the effectiveness of DPF by achieving superior performance. In addition, extensive analyses reveal that utilizing the multi-intent-free data can effectively enhance multi-intent SLU.
Large Language Models (LLM) can invoke a variety of tools and APIs to complete complex tasks. The computer, as the most powerful and universal tool, could potentially be controlled by a trained LLM agent. Powered by the computer, we can hopefully build a more generalized agent to assist humans in various daily digital works. In this paper, we construct an environment for a Vision Language Model (VLM) agent to interact with a real computer screen. Within this environment, the agent can observe screenshots and manipulate the Graphical User Interface (GUI) by outputting mouse and keyboard actions. We also design an automated control pipeline that includes planning, acting, and reflecting phases, guiding the agent to continuously interact with the environment and complete multi-step tasks. Additionally, we construct the ScreenAgent Dataset, which collects screenshots and action sequences when completing daily computer tasks. Finally, we train a model, ScreenAgent, which achieves comparable computer control capabilities to GPT-4V and demonstrated more precise UI positioning capabilities. Our attempts could inspire further research on building a generalist LLM agent. The code and more detailed information are at https://github.com/niuzaisheng/ScreenAgent.
Dialogue relation extraction (DRE) aims to extract relations between two arguments within a dialogue, which is more challenging than standard RE due to the higher person pronoun frequency and lower information density in dialogues. However, existing DRE methods still suffer from two serious issues: (1) hard to capture long and sparse multi-turn information, and (2) struggle to extract golden relations based on partial dialogues, which motivates us to discover more effective methods that can alleviate the above issues. We notice that the rise of large language models (LLMs) has sparked considerable interest in evaluating their performance across diverse tasks. To this end, we initially investigate the capabilities of different LLMs in DRE, considering both proprietary models and open-source models. Interestingly, we discover that LLMs significantly alleviate two issues in existing DRE methods. Generally, we have following findings: (1) scaling up model size substantially boosts the overall DRE performance and achieves exceptional results, tackling the difficulty of capturing long and sparse multi-turn information; (2) LLMs encounter with much smaller performance drop from entire dialogue setting to partial dialogue setting compared to existing methods; (3) LLMs deliver competitive or superior performances under both full-shot and few-shot settings compared to current state-of-the-art; (4) LLMs show modest performances on inverse relations but much stronger improvements on general relations, and they can handle dialogues of various lengths especially for longer sequences.
Designing natural language interfaces has historically required collecting supervised data to translate user requests into carefully designed intent representations. This requires enumerating and labeling a long tail of user requests, which is challenging. At the same time, large language models (LLMs) encode knowledge about goals and plans that can help conversational assistants interpret user requests requiring numerous steps to complete. We introduce an approach to handle complex-intent-bearing utterances from a user via a process of hierarchical natural language decomposition and interpretation. Our approach uses a pre-trained language model to decompose a complex utterance into a sequence of simpler natural language steps and interprets each step using the language-to-program model designed for the interface. To test our approach, we collect and release DeCU —a new NL-to-program benchmark to evaluate Decomposition of Complex Utterances. Experiments show that the proposed approach enables the interpretation of complex utterances with almost no complex training data, while outperforming standard few-shot prompting approaches.
The core of the dialogue system is to generate relevant, informative, and human-like responses based on extensive dialogue history. Recently, dialogue generation domain has seen mainstream adoption of large language models (LLMs), due to its powerful capability in generating utterances. However, there is a natural deficiency for such models, that is, inherent position bias, which may lead them to pay more attention to the nearby utterances instead of causally relevant ones, resulting in generating irrelevant and generic responses in long-term dialogue. To alleviate such problem, in this paper, we propose a novel method, named Causal Perception long-term Dialogue framework (CPD), which employs perturbation-based causal variable discovery method to extract casually relevant utterances from the dialogue history and enhances model causal perception during fine-tuning. Specifically, a local-position awareness method is proposed in CPD for inter-sentence position correlation elimination, which helps models extract causally relevant utterances based on perturbations. Then, a casual-perception fine-tuning strategy is also proposed, to enhance the capability of discovering the causal invariant factors, by differently perturbing causally relevant and non-casually relevant ones for response generation. Experimental results on two datasets prove that our proposed method can effectively alleviate the position bias for multiple LLMs and achieve significant progress compared with existing baselines.
End-to-end spoken language understanding (SLU) aims to directly capture the comprehensive semantics from the given spoken utterance without generating any transcript. Since the transcripts might not always be available, Textless SLU is attracting increasing attention, which could eliminate the need for transcripts but often does not perform as well as SLU models trained with transcripts. In this paper, we focus on the scenarios where the transcripts are not available and propose a framework GMA-SLU to generate more audios according to the labels. In order to alleviate the modality gap between text and audio, two language models are developed and discrete tokens are utilized as a bridge, where the first language model utilizes labels to generate semantic tokens and the second language model adopts these obtained semantic tokens and the acoustic tokens of source audios to generate the synthetic audios. All the experiments are conducted on the monolingual SLU dataset SLURP and the multilingual SLU dataset MINDS-14. Experimental results show that our method outperforms the previous best Textless End-to-end SLU models and can obtain the comparable performance with the models trained with the assistance of the corresponding transcripts.
Dialogue Cross-Enhanced Central Engagement Attention Model for Real-Time Engagement Estimation
PDF ↗Real-time engagement estimation has been an important research topic in human-computer interaction in recent years. The emergence of the NOvice eXpert Interaction (NOXI) dataset, enriched with frame-wise engagement annotations, has catalyzed a surge in research efforts in this domain. Existing feature sequence partitioning methods for ultra-long videos have encountered challenges including insufficient information utilization and repetitive inference. Moreover, those studies focus mainly on the target participants’ features without taking into account those of the interlocutor. To address these issues, we propose the center-based sliding window method to obtain feature subsequences. The core of these subsequences is modeled using our innovative Central Engagement Attention Model (CEAM). Additionally, we introduce the dialogue cross-enhanced module that effectively incorporates the interlocutor’s features via cross-attention. Our proposed method outperforms the current best model, achieving a substantial gain of 1.5% in coordination correlation coefficient (CCC) and establishing a new state-of-the-art result. Our source codes and model checkpoints are available at https://github.com/wujiekd/Dialogue-Cross-Enhanced-CEAM.
Human-AI interactivity is a critical aspect that reflects the usability of Multimodal Large Language Models (MLLMs). However, existing end-to-end MLLMs only allow users to interact with them through language instructions, leading to the limitation of the interactive accuracy and efficiency. In this study, we present precise referring instructions that utilize diverse reference representations such as points and boxes as referring prompts to refer to the special region. This enables MLLMs to focus on the region of interest and achieve finer-grained interaction. Based on precise referring instruction, we propose ChatSpot, a unified end-to-end MLLM that supports diverse forms of interactivity including mouse clicks, drag-and-drop, and drawing boxes, which provides a more flexible and seamless interactive experience. We also construct a multi-grained vision-language instruction-following dataset based on existing datasets and GPT-4 generating. Furthermore, we design a series of evaluation tasks to assess the effectiveness of region recognition and interaction. Experimental results showcase ChatSpot's promising performance. Project page: https://github.com/Ahnsun/ChatSpot.