Despite the remarkable capabilities of large language models (LLMs) in natural language understanding and reasoning, they often display undesirable behaviors, such as generating hallucinations and unfaithful reasoning. A prevalent strategy to mitigate these issues is the use of reflection, which refines responses through an iterative process. However, while promising, reflection heavily relies on high-quality external feedback and requires iterative multi-agent inference processes, thus hindering its practical application. In this paper, we propose Meta-Reflection, a novel feedback-free reflection mechanism that necessitates only a single inference pass without external feedback. Motivated by the human ability to remember and retrieve reflections from past experiences when encountering similar problems, Meta-Reflection integrates reflective insights into a codebook, allowing the historical insights to be stored, retrieved, and used to guide LLMs in problem-solving. To thoroughly investigate and evaluate the practicality of Meta-Reflection in real-world scenarios, we introduce an industrial e-commerce benchmark named E-commerce Customer Intent Detection (ECID). Extensive experiments conducted on both public datasets and the ECID benchmark highlight the effectiveness and efficiency of our proposed approach.
In this paper we present our submission for the NorSID Shared Task as part of the 2025 VarDial Workshop (Scherrer et al., 2025), consisting of three tasks: Intent Detection, Slot Filling and Dialect Identification, evaluated using data in different dialects of the Norwegian language. For Intent Detection and Slot Filling, we have fine-tuned a multitask model in a cross-lingual setting, to leverage the xSID dataset available in 17 languages. In the case of Dialect Identification, our final submission consists of a model fine-tuned on the provided development set, which has obtained the highest scores within our experiments. Our final results on the test set show that our models do not drop in performance compared to the development set, likely due to the domain-specificity of the dataset and the similar distribution of both subsets. Finally, we also report an in-depth analysis of the provided datasets and their artifacts, as well as other sets of experiments that have been carried out but did not yield the best results. Additionally, we present an analysis on the reasons why some methods have been more successful than others; mainly the impact of the combination of languages and domain-specificity of the training data on the results.
We explore the potential of pixel-based models for transfer learning from standard languages to dialects. These models convert text into images that are divided into patches, enabling a continuous vocabulary representation that proves especially useful for out-of-vocabulary words common in dialectal data. Using German as a case study, we compare the performance of pixel-based models to token-based models across various syntactic and semantic tasks. Our results show that pixel-based models outperform token-based models in part-of-speech tagging, dependency parsing and intent detection for zero-shot dialect evaluation by up to 26 percentage points in some scenarios, though not in Standard German. However, pixel-based models fall short in topic classification. These findings emphasize the potential of pixel-based models for handling dialectal data, though further research should be conducted to assess their effectiveness in various linguistic contexts.
Both encoder-only models (e.g., BERT, RoBERTa) and large language models (LLMs, e.g., Llama3) have been widely used for text classification tasks. However, there is a lack of systematic studies comparing the performance of encoder-based models and LLMs in text classification, particularly when fine-tuning is involved. This study employed a diverse range of models and methods, varying in size and architecture, and including both fine-tuned and pre-trained approaches. We first assessed the performances of these LLMs on the 20 Newsgroups (20NG) and MASSIVE datasets, comparing them to encoder-only RoBERTa models. Additionally, we explored the multi-task capabilities of both model types by combining multiple classification tasks, including intent detection and slot-filling, into a single model using data from both datasets. Our results indicate that fully fine-tuned Llama3-70B models outperform RoBERTa-large and other decoder LLMs across various classification tasks and datasets. Moreover, the consolidated multi-task fine-tuned LLMs matched the performance of dual-model setups in both tasks across both datasets. Overall, our study provides a comprehensive benchmark of encoder-only and LLM models on text classification tasks and demonstrates a method to combine two or more fully fine-tuned decoder LLMs for reduced latency and equivalent performance.
Intention recognition, or the ability to anticipate the actions of another agent, plays a vital role in the design and development of automated assistants that can support humans in their daily tasks. In particular, industrial settings pose interesting challenges that include potential distractions for a decision-maker as well as noisy or incomplete observations. In such a setting, a robotic assistant tasked with helping and supporting a human worker must interleave information gathering actions with proactive tasks of its own, an approach that has been referred to as active goal recognition. In this paper we describe a partially observable model for online intention recognition, show some preliminary experimental results and discuss some of the challenges present in this family of problems.
In today's digitally driven world, dialogue systems play a pivotal role in enhancing user interactions, from customer service to virtual assistants. In these dialogues, it is important to identify user's goals automatically to resolve their needs promptly. This has necessitated the integration of models that perform Intent Detection. However, users' intents are diverse and dynamic, making it challenging to maintain a fixed set of predefined intents. As a result, a more practical approach is to develop a model capable of identifying new intents as they emerge. We address the challenge of Intent Discovery, an area that has drawn significant attention in recent research efforts. Existing methods need to train on a substantial amount of data for correctly identifying new intents, demanding significant human effort. To overcome this, we introduce IntentGPT, a novel training-free method that effectively prompts Large Language Models (LLMs) such as GPT-4 to discover new intents with minimal labeled data. IntentGPT comprises an \textit{In-Context Prompt Generator}, which generates informative prompts for In-Context Learning, an \textit{Intent Predictor} for classifying and discovering user intents from utterances, and a \textit{Semantic Few-Shot Sampler} that selects relevant few-shot examples and a set of known intents to be injected into the prompt. Our experiments show that IntentGPT outperforms previous methods that require extensive domain-specific data and fine-tuning, in popular benchmarks, including CLINC and BANKING, among others.
In task-oriented dialogue systems, intent detection is crucial for interpreting user queries and providing appropriate responses. Existing research primarily addresses simple queries with a single intent, lacking effective systems for handling complex queries with multiple intents and extracting different intent spans. Additionally, there is a notable absence of multilingual, multi-intent datasets. This study addresses three critical tasks: extracting multiple intent spans from queries, detecting multiple intents, and developing a multi-lingual multi-label intent dataset. We introduce a novel multi-label multi-class intent detection dataset (MLMCID-dataset) curated from existing benchmark datasets. We also propose a pointer network-based architecture (MLMCID) to extract intent spans and detect multiple intents with coarse and fine-grained labels in the form of sextuplets. Comprehensive analysis demonstrates the superiority of our pointer network-based system over baseline approaches in terms of accuracy and F1-score across various datasets.
In this short paper we propose a data augmentation method for intent detection in zero-resource domains. Existing data augmentation methods rely on few labelled examples for each intent category, which can be expensive in settings with many possible intents. We use a two-stage approach: First, we generate utterances for intent labels using an open-source large language model in a zero-shot setting. Second, we develop a smaller sequence-to-sequence model (the Refiner), to improve the generated utterances. The Refiner is fine-tuned on seen domains and then applied to unseen domains. We evaluate our method by training an intent classifier on the generated data, and evaluating it on real (human) data. We find that the Refiner significantly improves the data utility and diversity over the zero-shot LLM baseline for unseen domains and over common baseline approaches. Our results indicate that a two-step approach of a generative LLM in zero-shot setting and a smaller sequence-to-sequence model can provide high-quality data for intent detection.
Intent detection is a critical component of task-oriented dialogue systems (TODS) which enables the identification of suitable actions to address user utterances at each dialog turn. Traditional approaches relied on computationally efficient supervised sentence transformer encoder models, which require substantial training data and struggle with out-of-scope (OOS) detection. The emergence of generative large language models (LLMs) with intrinsic world knowledge presents new opportunities to address these challenges. In this work, we adapt 7 SOTA LLMs using adaptive in-context learning and chain-of-thought prompting for intent detection, and compare their performance with contrastively fine-tuned sentence transformer (SetFit) models to highlight prediction quality and latency tradeoff. We propose a hybrid system using uncertainty based routing strategy to combine the two approaches that along with negative data augmentation results in achieving the best of both worlds ( i.e. within 2% of native LLM accuracy with 50% less latency). To better understand LLM OOS detection capabilities, we perform controlled experiments revealing that this capability is significantly influenced by the scope of intent labels and the size of the label space. We also introduce a two-step approach utilizing internal LLM representations, demonstrating empirical gains in OOS detection accuracy and F1-score by >5% for the Mistral-7B model.
Human-Robot Collaboration (HRC) is vital in Industry 4.0, using sensors, digital twins, collaborative robots (cobots), and intention-recognition models to have efficient manufacturing processes. However, Concept Drift is a significant challenge, where robots struggle to adapt to new environments. We address concept drift by integrating Adaptive Intelligence and self-labeling (SLB) to improve the resilience of intention-recognition in an HRC system. Our methodology begins with data collection using cameras and weight sensors, which is followed by annotation of intentions and state changes. Then we train various deep learning models with different preprocessing techniques for recognizing and predicting the intentions. Additionally, we developed a custom state detection algorithm for enhancing the accuracy of SLB, offering precise state-change definitions and timestamps to label intentions. Our results show that the MViT2 model with skeletal posture preprocessing achieves an accuracy of 83% on our data environment, compared to the 79% accuracy of MViT2 without skeleton posture extraction. Additionally, our SLB mechanism achieves a labeling accuracy of 91%, reducing a significant amount of time that would've been spent on manual annotation. Lastly, we observe swift scaling of model performance that combats concept drift by fine tuning on different increments of self-labeled data in a shifted domain that has key differences from the original training environment.. This study demonstrates the potential for rapid deployment of intelligent cobots in manufacturing through the steps shown in our methodology, paving a way for more adaptive and efficient HRC systems.
Deep learning models, such as those used for face recognition and attribute prediction, are susceptible to manipulations like adversarial noise and unintentional noise, including Gaussian and impulse noise. This paper introduces CIAI, a Class-Independent Adversarial Intent detection network built on a modified vision transformer with detection layers. CIAI employs a novel loss function that combines Maximum Mean Discrepancy and Center Loss to detect both intentional (adversarial attacks) and unintentional noise, regardless of the image class. It is trained in a multi-step fashion. We also introduce the aspect of intent during detection that can act as an added layer of security. We further showcase the performance of our proposed detector on CelebA, CelebA-HQ, LFW, AgeDB, and CIFAR-10 datasets. Our detector is able to detect both intentional (like FGSM, PGD, and DeepFool) and unintentional (like Gaussian and Salt & Pepper noises) perturbations.
Integrating machine learning (ML) into customer service chatbots enhances their ability to understand and respond to user queries, ultimately improving service performance. However, they may appear artificial to some users and affecting customer experience. Hence, meticulous evaluation of ML models for each pipeline component is crucial for optimizing performance, though differences in functionalities can lead to unfair comparisons. In this paper, we present a tailored experimental evaluation approach for goal-oriented customer service chatbots with pipeline architecture, focusing on three key components: Natural Language Understanding (NLU), dialogue management (DM), and Natural Language Generation (NLG). Our methodology emphasizes individual assessment to determine optimal ML models. Specifically, we focus on optimizing hyperparameters and evaluating candidate models for NLU (utilizing BERT and LSTM), DM (employing DQN and DDQN), and NLG (leveraging GPT-2 and DialoGPT). The results show that for the NLU component, BERT excelled in intent detection whereas LSTM was superior for slot filling. For the DM component, the DDQN model outperformed DQN by achieving fewer turns, higher rewards, as well as greater success rates. For NLG, the large language model GPT-2 surpassed DialoGPT in BLEU, METEOR, and ROUGE metrics. These findings aim to provide a benchmark for future research in developing and optimizing customer service chatbots, offering valuable insights into model performance and optimal hyperparameters.
Emotional intelligence in large language models (LLMs) is of great importance in Natural Language Processing. However, the previous research mainly focus on basic sentiment analysis tasks, such as emotion recognition, which is not enough to evaluate LLMs' overall emotional intelligence. Therefore, this paper presents a novel framework named EmotionQueen for evaluating the emotional intelligence of LLMs. The framework includes four distinctive tasks: Key Event Recognition, Mixed Event Recognition, Implicit Emotional Recognition, and Intention Recognition. LLMs are requested to recognize important event or implicit emotions and generate empathetic response. We also design two metrics to evaluate LLMs' capabilities in recognition and response for emotion-related statements. Experiments yield significant conclusions about LLMs' capabilities and limitations in emotion intelligence.
In the realm of task-oriented dialogue systems, a robust intent detection mechanism must effectively handle malformed utterances encountered in real-world scenarios. This study presents a novel fine-tuning framework for large language models (LLMs) aimed at enhancing in-distribution (ID) intent classification and out-of-distribution (OOD) intent detection, which utilizes semantic matching with prototypes derived from ID class names. By harnessing the highly distinguishable representations of LLMs, we construct semantic prototypes for each ID class using a diversity-grounded prompt tuning approach. We rigorously test our framework in a challenging OOD context, where ID and OOD classes are semantically close yet distinct, referred to as \emph{near} OOD detection. For a thorough assessment, we benchmark our method against the prevalent fine-tuning approaches. The experimental findings reveal that our method demonstrates superior performance in both few-shot ID intent classification and near-OOD intent detection tasks.
Although Large Language Models(LLMs) can generate coherent and contextually relevant text, they often struggle to recognise the intent behind the human user's query. Natural Language Understanding (NLU) models, however, interpret the purpose and key information of user's input to enable responsive interactions. Existing NLU models generally map individual utterances to a dual-level semantic frame, involving sentence-level intent and word-level slot labels. However, real-life conversations primarily consist of multi-turn conversations, involving the interpretation of complex and extended dialogues. Researchers encounter challenges addressing all facets of multi-turn dialogue conversations using a unified single NLU model. This paper introduces a novel approach, MIDAS, leveraging a multi-level intent, domain, and slot knowledge distillation for multi-turn NLU. To achieve this, we construct distinct teachers for varying levels of conversation knowledge, namely, sentence-level intent detection, word-level slot filling, and conversation-level domain classification. These teachers are then fine-tuned to acquire specific knowledge of their designated levels. A multi-teacher loss is proposed to facilitate the combination of these multi-level teachers, guiding a student model in multi-turn dialogue tasks. The experimental results demonstrate the efficacy of our model in improving the overall multi-turn conversation understanding, showcasing the potential for advancements in NLU models through the incorporation of multi-level dialogue knowledge distillation techniques.
Objective: This work explores use of a few-shot transfer learning method to train and implement a convolutional spiking neural network (CSNN) on a BrainChip Akida AKD1000 neuromorphic system-on-chip for developing individual-level, instead of traditionally used group-level, models using electroencephalographic data. Main Results: Efficacy of the above methodology to develop individual-specific braking intention predictive models by rapidly adapting the group-level model in as few as three training epochs while achieving at least 90% accuracy, true positive rate and true negative rate is presented. Further, results show the energy-efficiency of the neuromorphic hardware through a power reduction of over 97% with only a $1.3* increase in latency when using the Akida AKD1000 processor for network inference compared to an Intel Xeon central processing unit. Similar results were obtained in a subsequent ablation study using a subset of five out of 19 channels.
Intent detection is a text classification task whose aim is to recognize and label the semantics behind a users query. It plays a critical role in various business applications. The output of the intent detection module strongly conditions the behavior of the whole system. This sequence analysis task is mainly tackled using deep learning techniques. Despite the widespread use of these techniques, the internal mechanisms used by networks to solve the problem are poorly understood. Recent lines of work have analyzed the computational mechanisms learned by RNNs from a dynamical systems perspective. In this work, we investigate how different RNN architectures solve the SNIPS intent detection problem. Sentences injected into trained networks can be interpreted as trajectories traversing a hidden state space. This space is constrained to a low-dimensional manifold whose dimensionality is related to the embedding and hidden layer sizes. To generate predictions, RNN steers the trajectories towards concrete regions, spatially aligned with the output layer matrix rows directions. Underlying the system dynamics, an unexpected fixed point topology has been identified with a limited number of attractors. Our results provide new insights into the inner workings of networks that solve the intent detection task.
In this paper, we present our dzFinNlp team's contribution for intent detection in financial conversational agents, as part of the AraFinNLP shared task. We experimented with various models and feature configurations, including traditional machine learning methods like LinearSVC with TF-IDF, as well as deep learning models like Long Short-Term Memory (LSTM). Additionally, we explored the use of transformer-based models for this task. Our experiments show promising results, with our best model achieving a micro F1-score of 93.02% and 67.21% on the ArBanking77 dataset, in the development and test sets, respectively.
The expanding financial markets of the Arab world require sophisticated Arabic NLP tools. To address this need within the banking domain, the Arabic Financial NLP (AraFinNLP) shared task proposes two subtasks: (i) Multi-dialect Intent Detection and (ii) Cross-dialect Translation and Intent Preservation. This shared task uses the updated ArBanking77 dataset, which includes about 39k parallel queries in MSA and four dialects. Each query is labeled with one or more of a common 77 intents in the banking domain. These resources aim to foster the development of robust financial Arabic NLP, particularly in the areas of machine translation and banking chat-bots. A total of 45 unique teams registered for this shared task, with 11 of them actively participated in the test phase. Specifically, 11 teams participated in Subtask 1, while only 1 team participated in Subtask 2. The winning team of Subtask 1 achieved F1 score of 0.8773, and the only team submitted in Subtask 2 achieved a 1.667 BLEU score.
This paper presents a novel and comprehensive solution to enhance both the robustness and efficiency of question answering (QA) systems through supervised contrastive learning (SCL). Training a high-performance QA system has become straightforward with pre-trained language models, requiring only a small amount of data and simple fine-tuning. However, despite recent advances, existing QA systems still exhibit significant deficiencies in functionality and training efficiency. We address the functionality issue by defining four key tasks: user input intent classification, out-of-domain input detection, new intent discovery, and continual learning. We then leverage a unified SCL-based representation learning method to efficiently build an intra-class compact and inter-class scattered feature space, facilitating both known intent classification and unknown intent detection and discovery. Consequently, with minimal additional tuning on downstream tasks, our approach significantly improves model efficiency and achieves new state-of-the-art performance across all tasks.