Paweł Budzianowski

Iñigo Casanueva, Ivan Vulić, Georgios P. Spithourakis et al.

We present NLU++, a novel dataset for natural language understanding (NLU) in task-oriented dialogue (ToD) systems, with the aim to provide a much more challenging evaluation environment for dialogue NLU models, up to date with the current application and industry requirements. NLU++ is divided into two domains (BANKING and HOTELS) and brings several crucial improvements over current commonly used NLU datasets. 1) NLU++ provides fine-grained domain ontologies with a large set of challenging multi-intent sentences, introducing and validating the idea of intent modules that can be combined into complex intents that convey complex user goals, combined with finer-grained and thus more challenging slot sets. 2) The ontology is divided into domain-specific and generic (i.e., domain-universal) intent modules that overlap across domains, promoting cross-domain reusability of annotated examples. 3) The dataset design has been inspired by the problems observed in industrial ToD systems, and 4) it has been collected, filtered and carefully annotated by dialogue NLU experts, yielding high-quality annotated data. Finally, we benchmark a series of current state-of-the-art NLU models on NLU++; the results demonstrate the challenging nature of the dataset, especially in low-data regimes, the validity of `intent modularisation', and call for further research on ToD NLU.

Gabor Fuisz, Ivan Vulić, Samuel Gibbons et al.

Transformer-based pretrained language models (PLMs) offer unmatched performance across the majority of natural language understanding (NLU) tasks, including a body of question answering (QA) tasks. We hypothesize that improvements in QA methodology can also be directly exploited in dialog NLU; however, dialog tasks must be \textit{reformatted} into QA tasks. In particular, we focus on modeling and studying \textit{slot labeling} (SL), a crucial component of NLU for dialog, through the QA optics, aiming to improve both its performance and efficiency, and make it more effective and resilient to working with limited task data. To this end, we make a series of contributions: 1) We demonstrate how QA-tuned PLMs can be applied to the SL task, reaching new state-of-the-art performance, with large gains especially pronounced in such low-data regimes. 2) We propose to leverage contextual information, required to tackle ambiguous values, simply through natural language. 3) Efficiency and compactness of QA-oriented fine-tuning are boosted through the use of lightweight yet effective adapter modules. 4) Trading-off some of the quality of QA datasets for their size, we experiment with larger automatically generated QA datasets for QA-tuning, arriving at even higher performance. Finally, our analysis suggests that our novel QA-based slot labeling models, supported by the PLMs, reach a performance ceiling in high-data regimes, calling for more challenging and more nuanced benchmarks in future work.

Guangzhi Sun, Chao Zhang, Ivan Vulić et al.

Manually annotating fine-grained slot-value labels for task-oriented dialogue (ToD) systems is an expensive and time-consuming endeavour. This motivates research into slot-filling methods that operate with limited amounts of labelled data. Moreover, the majority of current work on ToD is based solely on text as the input modality, neglecting the additional challenges of imperfect automatic speech recognition (ASR) when working with spoken language. In this work, we propose a Knowledge-Aware Audio-Grounded generative slot-filling framework, termed KA2G, that focuses on few-shot and zero-shot slot filling for ToD with speech input. KA2G achieves robust and data-efficient slot filling for speech-based ToD by 1) framing it as a text generation task, 2) grounding text generation additionally in the audio modality, and 3) conditioning on available external knowledge (e.g. a predefined list of possible slot values). We show that combining both modalities within the KA2G framework improves the robustness against ASR errors. Further, the knowledge-aware slot-value generator in KA2G, implemented via a pointer generator mechanism, particularly benefits few-shot and zero-shot learning. Experiments, conducted on the standard speech-based single-turn SLURP dataset and a multi-turn dataset extracted from a commercial ToD system, display strong and consistent gains over prior work, especially in few-shot and zero-shot setups.

Georgios P. Spithourakis, Ivan Vulić, Michał Lis et al.

Knowledge-based authentication is crucial for task-oriented spoken dialogue systems that offer personalised and privacy-focused services. Such systems should be able to enrol (E), verify (V), and identify (I) new and recurring users based on their personal information, e.g. postcode, name, and date of birth. In this work, we formalise the three authentication tasks and their evaluation protocols, and we present EVI, a challenging spoken multilingual dataset with 5,506 dialogues in English, Polish, and French. Our proposed models set the first competitive benchmarks, explore the challenges of multilingual natural language processing of spoken dialogue, and set directions for future research.

Evgeniia Razumovskaia, Ivan Vulić, Pavle Marković et al.

Factuality is a crucial requirement in information seeking dialogue: the system should respond to the user's queries so that the responses are meaningful and aligned with the knowledge provided to the system. However, most modern large language models suffer from hallucinations, that is, they generate responses not supported by or contradicting the knowledge source. To mitigate the issue and increase faithfulness of information-seeking dialogue systems, we introduce BeInfo, a simple yet effective method that applies behavioural tuning to aid information-seeking dialogue. Relying on three standard datasets, we show that models tuned with BeInfo} become considerably more faithful to the knowledge source both for datasets and domains seen during BeInfo-tuning, as well as on unseen domains, when applied in a zero-shot manner. In addition, we show that the models with 3B parameters (e.g., Flan-T5) tuned with BeInfo demonstrate strong performance on data from real `production' conversations and outperform GPT4 when tuned on a limited amount of such realistic in-domain dialogues.

Paweł Budzianowski, Wesley Maa, Matthew Freed et al.

Vision-Language Models (VLMs) have emerged as a promising approach to address the data scarcity challenge in robotics, enabling the development of generalizable visuomotor control policies. While models like OpenVLA showcase the potential of this paradigm, deploying large-scale VLMs on resource-constrained mobile manipulation systems remains a significant hurdle. This paper introduces Edge VLA (EVLA), a novel approach designed to significantly enhance the inference speed of Vision-Language-Action (VLA) models. EVLA maintains the representational power of these models while enabling real-time performance on edge devices. We achieve this through two key innovations: 1) Eliminating the autoregressive requirement for end-effector position prediction, leading to a 7x speedup in inference, and 2) Leveraging the efficiency of Small Language Models (SLMs), demonstrating comparable training performance to larger models with significantly reduced computational demands. Our early results demonstrate that EVLA achieves comparable training characteristics to OpenVLA while offering substantial gains in inference speed and memory efficiency. We release our model checkpoints and training \href{https://github.com/kscalelabs/evla }{codebase} to foster further research.

Gracjan Góral, Emilia Wiśnios, Piotr Sankowski et al.

This work introduces a novel framework for evaluating LLMs' capacity to balance instruction-following with critical reasoning when presented with multiple-choice questions containing no valid answers. Through systematic evaluation across arithmetic, domain-specific knowledge, and high-stakes medical decision tasks, we demonstrate that post-training aligned models often default to selecting invalid options, while base models exhibit improved refusal capabilities that scale with model size. Our analysis reveals that alignment techniques, though intended to enhance helpfulness, can inadvertently impair models' reflective judgment--the ability to override default behaviors when faced with invalid options. We additionally conduct a parallel human study showing similar instruction-following biases, with implications for how these biases may propagate through human feedback datasets used in alignment. We provide extensive ablation studies examining the impact of model size, training techniques, and prompt engineering. Our findings highlight fundamental tensions between alignment optimization and preservation of critical reasoning capabilities, with important implications for developing more robust AI systems for real-world deployment.

Paweł Budzianowski, Emilia Wiśnios, Gracjan Góral et al.

Data scarcity remains one of the most limiting factors in driving progress in robotics. However, the amount of available robotics data in the wild is growing exponentially, creating new opportunities for large-scale data utilization. Reliable temporal task completion prediction could help automatically annotate and curate this data at scale. The Generative Value Learning (GVL) approach was recently proposed, leveraging the knowledge embedded in vision-language models (VLMs) to predict task progress from visual observations. Building upon GVL, we propose OpenGVL, a comprehensive benchmark for estimating task progress across diverse challenging manipulation tasks involving both robotic and human embodiments. We evaluate the capabilities of publicly available open-source foundation models, showing that open-source model families significantly underperform closed-source counterparts, achieving only approximately $70\%$ of their performance on temporal progress prediction tasks. Furthermore, we demonstrate how OpenGVL can serve as a practical tool for automated data curation and filtering, enabling efficient quality assessment of large-scale robotics datasets. We release the benchmark along with the complete codebase at \href{github.com/budzianowski/opengvl}{OpenGVL}.

Paweł Budzianowski, Tsung-Hsien Wen, Bo-Hsiang Tseng et al.

Even though machine learning has become the major scene in dialogue research community, the real breakthrough has been blocked by the scale of data available. To address this fundamental obstacle, we introduce the Multi-Domain Wizard-of-Oz dataset (MultiWOZ), a fully-labeled collection of human-human written conversations spanning over multiple domains and topics. At a size of $10$k dialogues, it is at least one order of magnitude larger than all previous annotated task-oriented corpora. The contribution of this work apart from the open-sourced dataset labelled with dialogue belief states and dialogue actions is two-fold: firstly, a detailed description of the data collection procedure along with a summary of data structure and analysis is provided. The proposed data-collection pipeline is entirely based on crowd-sourcing without the need of hiring professional annotators; secondly, a set of benchmark results of belief tracking, dialogue act and response generation is reported, which shows the usability of the data and sets a baseline for future studies.

Paweł Budzianowski, Taras Sereda, Tomasz Cichy et al.

In recent years, speech generation has seen remarkable progress, now achieving one-shot generation capability that is often virtually indistinguishable from real human voice. Integrating such advancements in speech generation with large language models might revolutionize a wide range of applications. However, certain applications, such as assistive conversational systems, require natural and conversational speech generation tools that also operate efficiently in real time. Current state-of-the-art models like VALL-E and SoundStorm, powered by hierarchical neural audio codecs, require large neural components and extensive training data to work well. In contrast, MQTTS aims to build more compact conversational TTS models while capitalizing on smaller-scale real-life conversational speech data. However, its autoregressive nature yields high inference latency and thus limits its real-time usage. In order to mitigate the current limitations of the state-of-the-art TTS models while capitalizing on their strengths, in this work we introduce the Pheme model series that 1) offers compact yet high-performing models, 2) allows for parallel speech generation of 3) natural conversational speech, and 4) it can be trained efficiently on smaller-scale conversational data, cutting data demands by more than 10x but still matching the quality of the autoregressive TTS models. We also show that through simple teacher-student distillation we can meet significant improvements in voice quality for single-speaker setups on top of pretrained Pheme checkpoints, relying solely on synthetic speech generated by much larger teacher models. Audio samples and pretrained models are available online.

Ivan Vulić, Pei-Hao Su, Sam Coope et al.

Transformer-based language models (LMs) pretrained on large text collections are proven to store a wealth of semantic knowledge. However, 1) they are not effective as sentence encoders when used off-the-shelf, and 2) thus typically lag behind conversationally pretrained (e.g., via response selection) encoders on conversational tasks such as intent detection (ID). In this work, we propose ConvFiT, a simple and efficient two-stage procedure which turns any pretrained LM into a universal conversational encoder (after Stage 1 ConvFiT-ing) and task-specialised sentence encoder (after Stage 2). We demonstrate that 1) full-blown conversational pretraining is not required, and that LMs can be quickly transformed into effective conversational encoders with much smaller amounts of unannotated data; 2) pretrained LMs can be fine-tuned into task-specialised sentence encoders, optimised for the fine-grained semantics of a particular task. Consequently, such specialised sentence encoders allow for treating ID as a simple semantic similarity task based on interpretable nearest neighbours retrieval. We validate the robustness and versatility of the ConvFiT framework with such similarity-based inference on the standard ID evaluation sets: ConvFiT-ed LMs achieve state-of-the-art ID performance across the board, with particular gains in the most challenging, few-shot setups.

Bo-Hsiang Tseng, Marek Rei, Paweł Budzianowski et al.

Dialogue systems benefit greatly from optimizing on detailed annotations, such as transcribed utterances, internal dialogue state representations and dialogue act labels. However, collecting these annotations is expensive and time-consuming, holding back development in the area of dialogue modelling. In this paper, we investigate semi-supervised learning methods that are able to reduce the amount of required intermediate labelling. We find that by leveraging un-annotated data instead, the amount of turn-level annotations of dialogue state can be significantly reduced when building a neural dialogue system. Our analysis on the MultiWOZ corpus, covering a range of domains and topics, finds that annotations can be reduced by up to 30\% while maintaining equivalent system performance. We also describe and evaluate the first end-to-end dialogue model created for the MultiWOZ corpus.

Bo-Hsiang Tseng, Paweł Budzianowski, Yen-Chen Wu et al.

Domain adaptation in natural language generation (NLG) remains challenging because of the high complexity of input semantics across domains and limited data of a target domain. This is particularly the case for dialogue systems, where we want to be able to seamlessly include new domains into the conversation. Therefore, it is crucial for generation models to share knowledge across domains for the effective adaptation from one domain to another. In this study, we exploit a tree-structured semantic encoder to capture the internal structure of complex semantic representations required for multi-domain dialogues in order to facilitate knowledge sharing across domains. In addition, a layer-wise attention mechanism between the tree encoder and the decoder is adopted to further improve the model's capability. The automatic evaluation results show that our model outperforms previous methods in terms of the BLEU score and the slot error rate, in particular when the adaptation data is limited. In subjective evaluation, human judges tend to prefer the sentences generated by our model, rating them more highly on informativeness and naturalness than other systems.

Joachim Bingel, Victor Petrén Bach Hansen, Ana Valeria Gonzalez et al.

Task oriented dialogue systems rely heavily on specialized dialogue state tracking (DST) modules for dynamically predicting user intent throughout the conversation. State-of-the-art DST models are typically trained in a supervised manner from manual annotations at the turn level. However, these annotations are costly to obtain, which makes it difficult to create accurate dialogue systems for new domains. To address these limitations, we propose a method, based on reinforcement learning, for transferring DST models to new domains without turn-level supervision. Across several domains, our experiments show that this method quickly adapts off-the-shelf models to new domains and performs on par with models trained with turn-level supervision. We also show our method can improve models trained using turn-level supervision by subsequent fine-tuning optimization toward dialog-level rewards.

Matthew Henderson, Ivan Vulić, Iñigo Casanueva et al.

We present PolyResponse, a conversational search engine that supports task-oriented dialogue. It is a retrieval-based approach that bypasses the complex multi-component design of traditional task-oriented dialogue systems and the use of explicit semantics in the form of task-specific ontologies. The PolyResponse engine is trained on hundreds of millions of examples extracted from real conversations: it learns what responses are appropriate in different conversational contexts. It then ranks a large index of text and visual responses according to their similarity to the given context, and narrows down the list of relevant entities during the multi-turn conversation. We introduce a restaurant search and booking system powered by the PolyResponse engine, currently available in 8 different languages.

Paweł Budzianowski, Ivan Vulić

Data scarcity is a long-standing and crucial challenge that hinders quick development of task-oriented dialogue systems across multiple domains: task-oriented dialogue models are expected to learn grammar, syntax, dialogue reasoning, decision making, and language generation from absurdly small amounts of task-specific data. In this paper, we demonstrate that recent progress in language modeling pre-training and transfer learning shows promise to overcome this problem. We propose a task-oriented dialogue model that operates solely on text input: it effectively bypasses explicit policy and language generation modules. Building on top of the TransferTransfo framework (Wolf et al., 2019) and generative model pre-training (Radford et al., 2019), we validate the approach on complex multi-domain task-oriented dialogues from the MultiWOZ dataset. Our automatic and human evaluations show that the proposed model is on par with a strong task-specific neural baseline. In the long run, our approach holds promise to mitigate the data scarcity problem, and to support the construction of more engaging and more eloquent task-oriented conversational agents.

Matthew Henderson, Ivan Vulić, Daniela Gerz et al.

Despite their popularity in the chatbot literature, retrieval-based models have had modest impact on task-oriented dialogue systems, with the main obstacle to their application being the low-data regime of most task-oriented dialogue tasks. Inspired by the recent success of pretraining in language modelling, we propose an effective method for deploying response selection in task-oriented dialogue. To train response selection models for task-oriented dialogue tasks, we propose a novel method which: 1) pretrains the response selection model on large general-domain conversational corpora; and then 2) fine-tunes the pretrained model for the target dialogue domain, relying only on the small in-domain dataset to capture the nuances of the given dialogue domain. Our evaluation on six diverse application domains, ranging from e-commerce to banking, demonstrates the effectiveness of the proposed training method.

Matthew Henderson, Paweł Budzianowski, Iñigo Casanueva et al.

Progress in Machine Learning is often driven by the availability of large datasets, and consistent evaluation metrics for comparing modeling approaches. To this end, we present a repository of conversational datasets consisting of hundreds of millions of examples, and a standardised evaluation procedure for conversational response selection models using '1-of-100 accuracy'. The repository contains scripts that allow researchers to reproduce the standard datasets, or to adapt the pre-processing and data filtering steps to their needs. We introduce and evaluate several competitive baselines for conversational response selection, whose implementations are shared in the repository, as well as a neural encoder model that is trained on the entire training set.

Osman Ramadan, Paweł Budzianowski, Milica Gašić

Robust dialogue belief tracking is a key component in maintaining good quality dialogue systems. The tasks that dialogue systems are trying to solve are becoming increasingly complex, requiring scalability to multi domain, semantically rich dialogues. However, most current approaches have difficulty scaling up with domains because of the dependency of the model parameters on the dialogue ontology. In this paper, a novel approach is introduced that fully utilizes semantic similarity between dialogue utterances and the ontology terms, allowing the information to be shared across domains. The evaluation is performed on a recently collected multi-domain dialogues dataset, one order of magnitude larger than currently available corpora. Our model demonstrates great capability in handling multi-domain dialogues, simultaneously outperforming existing state-of-the-art models in single-domain dialogue tracking tasks.

Iñigo Casanueva, Paweł Budzianowski, Pei-Hao Su et al.

Reinforcement learning (RL) is a promising approach to solve dialogue policy optimisation. Traditional RL algorithms, however, fail to scale to large domains due to the curse of dimensionality. We propose a novel Dialogue Management architecture, based on Feudal RL, which decomposes the decision into two steps; a first step where a master policy selects a subset of primitive actions, and a second step where a primitive action is chosen from the selected subset. The structural information included in the domain ontology is used to abstract the dialogue state space, taking the decisions at each step using different parts of the abstracted state. This, combined with an information sharing mechanism between slots, increases the scalability to large domains. We show that an implementation of this approach, based on Deep-Q Networks, significantly outperforms previous state of the art in several dialogue domains and environments, without the need of any additional reward signal.

Gellért Weisz, Paweł Budzianowski, Pei-Hao Su et al.

In spoken dialogue systems, we aim to deploy artificial intelligence to build automated dialogue agents that can converse with humans. A part of this effort is the policy optimisation task, which attempts to find a policy describing how to respond to humans, in the form of a function taking the current state of the dialogue and returning the response of the system. In this paper, we investigate deep reinforcement learning approaches to solve this problem. Particular attention is given to actor-critic methods, off-policy reinforcement learning with experience replay, and various methods aimed at reducing the bias and variance of estimators. When combined, these methods result in the previously proposed ACER algorithm that gave competitive results in gaming environments. These environments however are fully observable and have a relatively small action set so in this paper we examine the application of ACER to dialogue policy optimisation. We show that this method beats the current state-of-the-art in deep learning approaches for spoken dialogue systems. This not only leads to a more sample efficient algorithm that can train faster, but also allows us to apply the algorithm in more difficult environments than before. We thus experiment with learning in a very large action space, which has two orders of magnitude more actions than previously considered. We find that ACER trains significantly faster than the current state-of-the-art.

Christopher Tegho, Paweł Budzianowski, Milica Gašić

In statistical dialogue management, the dialogue manager learns a policy that maps a belief state to an action for the system to perform. Efficient exploration is key to successful policy optimisation. Current deep reinforcement learning methods are very promising but rely on epsilon-greedy exploration, thus subjecting the user to a random choice of action during learning. Alternative approaches such as Gaussian Process SARSA (GPSARSA) estimate uncertainties and are sample efficient, leading to better user experience, but on the expense of a greater computational complexity. This paper examines approaches to extract uncertainty estimates from deep Q-networks (DQN) in the context of dialogue management. We perform an extensive benchmark of deep Bayesian methods to extract uncertainty estimates, namely Bayes-By-Backprop, dropout, its concrete variation, bootstrapped ensemble and alpha-divergences, combining it with DQN algorithm.

Iñigo Casanueva, Paweł Budzianowski, Pei-Hao Su et al.

Dialogue assistants are rapidly becoming an indispensable daily aid. To avoid the significant effort needed to hand-craft the required dialogue flow, the Dialogue Management (DM) module can be cast as a continuous Markov Decision Process (MDP) and trained through Reinforcement Learning (RL). Several RL models have been investigated over recent years. However, the lack of a common benchmarking framework makes it difficult to perform a fair comparison between different models and their capability to generalise to different environments. Therefore, this paper proposes a set of challenging simulated environments for dialogue model development and evaluation. To provide some baselines, we investigate a number of representative parametric algorithms, namely deep reinforcement learning algorithms - DQN, A2C and Natural Actor-Critic and compare them to a non-parametric model, GP-SARSA. Both the environments and policy models are implemented using the publicly available PyDial toolkit and released on-line, in order to establish a testbed framework for further experiments and to facilitate experimental reproducibility.

Stefan Ultes, Paweł Budzianowski, Iñigo Casanueva et al.

Reinforcement learning is widely used for dialogue policy optimization where the reward function often consists of more than one component, e.g., the dialogue success and the dialogue length. In this work, we propose a structured method for finding a good balance between these components by searching for the optimal reward component weighting. To render this search feasible, we use multi-objective reinforcement learning to significantly reduce the number of training dialogues required. We apply our proposed method to find optimized component weights for six domains and compare them to a default baseline.

Paweł Budzianowski, Stefan Ultes, Pei-Hao Su et al.

Human conversation is inherently complex, often spanning many different topics/domains. This makes policy learning for dialogue systems very challenging. Standard flat reinforcement learning methods do not provide an efficient framework for modelling such dialogues. In this paper, we focus on the under-explored problem of multi-domain dialogue management. First, we propose a new method for hierarchical reinforcement learning using the option framework. Next, we show that the proposed architecture learns faster and arrives at a better policy than the existing flat ones do. Moreover, we show how pretrained policies can be adapted to more complex systems with an additional set of new actions. In doing that, we show that our approach has the potential to facilitate policy optimisation for more sophisticated multi-domain dialogue systems.