Prompt compression condenses contexts while maintaining their informativeness for different usage scenarios. It not only shortens the inference time and reduces computational costs during the usage of large language models, but also lowers expenses when using closed-source models. In a preliminary study, we discover that when instructing language models to compress prompts, different compression styles (e.g., extractive or abstractive) impact performance of compressed prompts on downstream tasks. Building on this insight, we propose Style-Compress, a lightweight framework that adapts a smaller language model to compress prompts for a larger model on a new task without additional training. Our approach iteratively generates and selects effective compressed prompts as task-specific demonstrations through style variation and in-context learning, enabling smaller models to act as efficient compressors with task-specific examples. Style-Compress outperforms two baseline compression models in four tasks: original prompt reconstruction, text summarization, multi-hop QA, and CoT reasoning. In addition, with only 10 samples and 100 queries for adaptation, prompts compressed by Style-Compress achieve performance on par with or better than original prompts at a compression ratio of 0.25 or 0.5.
Multimodal Chain of Thought (MCoT) is a popular prompting strategy for improving the performance of multimodal large language models (MLLMs) across a range of complex reasoning tasks. Despite its popularity, there is a notable absence of automated methods for evaluating the quality of reasoning steps in MCoT. To address this gap, we propose Multimodal Chain-of-Thought Evaluation (MiCEval), a framework designed to assess the correctness of reasoning chains by evaluating the quality of both the description and each reasoning step. The evaluation of the description component focuses on the accuracy of the image descriptions, while the reasoning step evaluates the quality of each step as it is conditionally generated based on the preceding steps. MiCEval is built upon a fine-grained dataset with annotations that rate each step according to correctness, relevance, and informativeness. Extensive experiments on four state-of-the-art MLLMs show that step-wise evaluations using MiCEval align more closely with human judgments compared to existing methods based on cosine similarity or fine-tuning approaches. MiCEval datasets and code can be found in https://github.com/alenai97/MiCEval.
In many application domains, the proliferation of sensors and devices is generating vast volumes of data, imposing significant pressure on existing data analysis and data mining techniques. Nevertheless, an increase in data volume does not inherently imply an increase in informational content, as a substantial portion may be redundant or represent noise. This challenge is particularly evident in the deep learning domain, where the utility of additional data is contingent on its informativeness. In the absence of such, larger datasets merely exacerbate the computational cost and complexity of the learning process. To address these challenges, we propose RAZOR, a novel instance selection technique designed to extract a significantly smaller yet sufficiently informative subset from a larger set of instances without compromising the learning process. RAZOR has been specifically engineered to be robust, efficient, and scalable, making it suitable for large-scale datasets. Unlike many techniques in the literature, RAZOR is capable of operating in both supervised and unsupervised settings. Experimental results demonstrate that RAZOR outperforms recent state-of-the-art techniques in terms of both effectiveness and efficiency.
Meeting summarization is crucial in digital communication, but existing solutions struggle with salience identification to generate personalized, workable summaries, and context understanding to fully comprehend the meetings' content. Previous attempts to address these issues by considering related supplementary resources (e.g., presentation slides) alongside transcripts are hindered by models' limited context sizes and handling the additional complexities of the multi-source tasks, such as identifying relevant information in additional files and seamlessly aligning it with the meeting content. This work explores multi-source meeting summarization considering supplementary materials through a three-stage large language model approach: identifying transcript passages needing additional context, inferring relevant details from supplementary materials and inserting them into the transcript, and generating a summary from this enriched transcript. Our multi-source approach enhances model understanding, increasing summary relevance by ~9% and producing more content-rich outputs. We introduce a personalization protocol that extracts participant characteristics and tailors summaries accordingly, improving informativeness by ~10%. This work further provides insights on performance-cost trade-offs across four leading model families, including edge-device capable options. Our approach can be extended to similar complex generative tasks benefitting from additional resources and personalization, such as dialogue systems and action planning.
Fairness metrics are a core tool in the fair machine learning literature (FairML), used to determine that ML models are, in some sense, "fair". Real-world data, however, are typically plagued by various measurement biases and other violated assumptions, which can render fairness assessments meaningless. We adapt tools from causal sensitivity analysis to the FairML context, providing a general framework which (1) accommodates effectively any combination of fairness metric and bias that can be posed in the "oblivious setting"; (2) allows researchers to investigate combinations of biases, resulting in non-linear sensitivity; and (3) enables flexible encoding of domain-specific constraints and assumptions. Employing this framework, we analyze the sensitivity of the most common parity metrics under 3 varieties of classifier across 14 canonical fairness datasets. Our analysis reveals the striking fragility of fairness assessments to even minor dataset biases. We show that causal sensitivity analysis provides a powerful and necessary toolkit for gauging the informativeness of parity metric evaluations. Our repository is available here: https://github.com/Jakefawkes/fragile_fair.
Knowledge graph-grounded dialog generation requires retrieving a dialog-relevant subgraph from the given knowledge base graph and integrating it with the dialog history. Previous works typically represent the graph using an external encoder, such as graph neural networks, and retrieve relevant triplets based on the similarity between single-vector representations of triplets and the dialog history. However, these external encoders fail to leverage the rich knowledge of pretrained language models, and the retrieval process is also suboptimal due to the information bottleneck caused by the single-vector abstraction of the dialog history. In this work, we propose Dialog generation with Generative Subgraph Retrieval (DialogGSR), which retrieves relevant knowledge subgraphs by directly generating their token sequences on top of language models. For effective generative subgraph retrieval, we introduce two key methods: (i) structure-aware knowledge graph linearization with self-supervised graph-specific tokens and (ii) graph-constrained decoding utilizing graph structural proximity-based entity informativeness scores for valid and relevant generative retrieval. DialogGSR achieves state-of-the-art performance in knowledge graph-grounded dialog generation, as demonstrated on OpenDialKG and KOMODIS datasets.
Human lexicons contain many different words that speakers can use to refer to the same object, e.g., "purple" or "magenta" for the same shade of color. On the one hand, studies on language use have explored how speakers adapt their referring expressions to successfully communicate in context, without focusing on properties of the lexical system. On the other hand, studies in language evolution have discussed how competing pressures for informativeness and simplicity shape lexical systems, without tackling in-context communication. We aim at bridging the gap between these traditions, and explore why a soft mapping between referents and words is a good solution for communication, by taking into account both in-context communication and the structure of the lexicon. We propose a simple measure of informativeness for words and lexical systems, grounded in a visual space, and analyze color naming data for English and Mandarin Chinese. We conclude that optimal lexical systems are those where multiple words can apply to the same referent, conveying different amounts of information. Such systems allow speakers to maximize communication accuracy and minimize the amount of information they convey when communicating about referents in contexts.
In knowledge graph embedding, aside from positive triplets (ie: facts in the knowledge graph), the negative triplets used for training also have a direct influence on the model performance. In reality, since knowledge graphs are sparse and incomplete, negative triplets often lack explicit labels, and thus they are often obtained from various sampling strategies (eg: randomly replacing an entity in a positive triplet). An ideal sampled negative triplet should be informative enough to help the model train better. However, existing methods often ignore diversity and adaptiveness in their sampling process, which harms the informativeness of negative triplets. As such, we propose a generative adversarial approach called Diversified and Adaptive Negative Sampling DANS on knowledge graphs. DANS is equipped with a two-way generator that generates more diverse negative triplets through two pathways, and an adaptive mechanism that produces more fine-grained examples by localizing the global generator for different entities and relations. On the one hand, the two-way generator increase the overall informativeness with more diverse negative examples; on the other hand, the adaptive mechanism increases the individual sample-wise informativeness with more fine-grained sampling. Finally, we evaluate the performance of DANS on three benchmark knowledge graphs to demonstrate its effectiveness through quantitative and qualitative experiments.
In ad text generation (ATG), desirable ad text is both faithful and informative. That is, it should be faithful to the input document, while at the same time containing important information that appeals to potential customers. The existing evaluation data, CAMERA (arXiv:2309.12030), is suitable for evaluating informativeness, as it consists of reference ad texts created by ad creators. However, these references often include information unfaithful to the input, which is a notable obstacle in promoting ATG research. In this study, we collaborate with in-house ad creators to refine the CAMERA references and develop an alternative ATG evaluation dataset called FaithCAMERA, in which the faithfulness of references is guaranteed. Using FaithCAMERA, we can evaluate how well existing methods for improving faithfulness can generate informative ad text while maintaining faithfulness. Our experiments show that removing training data that contains unfaithful entities improves the faithfulness and informativeness at the entity level, but decreases both at the sentence level. This result suggests that for future ATG research, it is essential not only to scale the training data but also to ensure their faithfulness. Our dataset will be publicly available.
We present a study of LLMs' performance in generating and rating example sentences for bilingual dictionaries across languages with varying resource levels: French (high-resource), Indonesian (mid-resource), and Tetun (low-resource), with English as the target language. We evaluate the quality of LLM-generated examples against the GDEX (Good Dictionary EXample) criteria: typicality, informativeness, and intelligibility. Our findings reveal that while LLMs can generate reasonably good dictionary examples, their performance degrades significantly for lower-resourced languages. We also observe high variability in human preferences for example quality, reflected in low inter-annotator agreement rates. To address this, we demonstrate that in-context learning can successfully align LLMs with individual annotator preferences. Additionally, we explore the use of pre-trained language models for automated rating of examples, finding that sentence perplexity serves as a good proxy for typicality and intelligibility in higher-resourced languages. Our study also contributes a novel dataset of 600 ratings for LLM-generated sentence pairs, and provides insights into the potential of LLMs in reducing the cost of lexicographic work, particularly for low-resource languages.
As Large Language Models (LLMs) grow increasingly powerful, ensuring their safety and alignment with human values remains a critical challenge. Ideally, LLMs should provide informative responses while avoiding the disclosure of harmful or sensitive information. However, current alignment approaches, which rely heavily on refusal strategies, such as training models to completely reject harmful prompts or applying coarse filters are limited by their binary nature. These methods either fully deny access to information or grant it without sufficient nuance, leading to overly cautious responses or failures to detect subtle harmful content. For example, LLMs may refuse to provide basic, public information about medication due to misuse concerns. Moreover, these refusal-based methods struggle to handle mixed-content scenarios and lack the ability to adapt to context-dependent sensitivities, which can result in over-censorship of benign content. To overcome these challenges, we introduce HiddenGuard, a novel framework for fine-grained, safe generation in LLMs. HiddenGuard incorporates Prism (rePresentation Router for In-Stream Moderation), which operates alongside the LLM to enable real-time, token-level detection and redaction of harmful content by leveraging intermediate hidden states. This fine-grained approach allows for more nuanced, context-aware moderation, enabling the model to generate informative responses while selectively redacting or replacing sensitive information, rather than outright refusal. We also contribute a comprehensive dataset with token-level fine-grained annotations of potentially harmful information across diverse contexts. Our experiments demonstrate that HiddenGuard achieves over 90% in F1 score for detecting and redacting harmful content while preserving the overall utility and informativeness of the model's responses.
Multi-fidelity Bayesian Optimization (MFBO) is a promising framework to speed up materials and molecular discovery as sources of information of different accuracies are at hand at increasing cost. Despite its potential use in chemical tasks, there is a lack of systematic evaluation of the many parameters playing a role in MFBO. In this work, we provide guidelines and recommendations to decide when to use MFBO in experimental settings. We investigate MFBO methods applied to molecules and materials problems. First, we test two different families of acquisition functions in two synthetic problems and study the effect of the informativeness and cost of the approximate function. We use our implementation and guidelines to benchmark three real discovery problems and compare them against their single-fidelity counterparts. Our results may help guide future efforts to implement MFBO as a routine tool in the chemical sciences.
Empathetic and coherent responses are critical in auto-mated chatbot-facilitated psychotherapy. This study addresses the challenge of enhancing the emotional and contextual understanding of large language models (LLMs) in psychiatric applications. We introduce Emotion-Aware Embedding Fusion, a novel framework integrating hierarchical fusion and attention mechanisms to prioritize semantic and emotional features in therapy transcripts. Our approach combines multiple emotion lexicons, including NRC Emotion Lexicon, VADER, WordNet, and SentiWordNet, with state-of-the-art LLMs such as Flan-T5, LLAMA 2, DeepSeek-R1, and ChatGPT 4. Therapy session transcripts, comprising over 2,000 samples are segmented into hierarchical levels (word, sentence, and session) using neural networks, while hierarchical fusion combines these features with pooling techniques to refine emotional representations. Atten-tion mechanisms, including multi-head self-attention and cross-attention, further prioritize emotional and contextual features, enabling temporal modeling of emotion-al shifts across sessions. The processed embeddings, computed using BERT, GPT-3, and RoBERTa are stored in the Facebook AI similarity search vector database, which enables efficient similarity search and clustering across dense vector spaces. Upon user queries, relevant segments are retrieved and provided as context to LLMs, enhancing their ability to generate empathetic and con-textually relevant responses. The proposed framework is evaluated across multiple practical use cases to demonstrate real-world applicability, including AI-driven therapy chatbots. The system can be integrated into existing mental health platforms to generate personalized responses based on retrieved therapy session data.
In digital healthcare, large language models (LLMs) have primarily been utilized to enhance question-answering capabilities and improve patient interactions. However, effective patient care necessitates LLM chains that can actively gather information by posing relevant questions. This paper presents HealthQ, a novel framework designed to evaluate the questioning capabilities of LLM healthcare chains. We implemented several LLM chains, including Retrieval-Augmented Generation (RAG), Chain of Thought (CoT), and reflective chains, and introduced an LLM judge to assess the relevance and informativeness of the generated questions. To validate HealthQ, we employed traditional Natural Language Processing (NLP) metrics such as Recall-Oriented Understudy for Gisting Evaluation (ROUGE) and Named Entity Recognition (NER)-based set comparison, and constructed two custom datasets from public medical note datasets, ChatDoctor and MTS-Dialog. Our contributions are threefold: we provide the first comprehensive study on the questioning capabilities of LLMs in healthcare conversations, develop a novel dataset generation pipeline, and propose a detailed evaluation methodology.
Automatic text summarization in Nepali language is an unexplored area in natural language processing (NLP). Although considerable research has been dedicated to extractive summarization, the area of abstractive summarization, especially for low-resource languages such as Nepali, remains largely unexplored. This study explores the use of multilingual transformer models, specifically mBART and mT5, for generating headlines for Nepali news articles through abstractive summarization. The research addresses key challenges associated with summarizing texts in Nepali by first creating a summarization dataset through web scraping from various Nepali news portals. These multilingual models were then fine-tuned using different strategies. The performance of the fine-tuned models were then assessed using ROUGE scores and human evaluation to ensure the generated summaries were coherent and conveyed the original meaning. During the human evaluation, the participants were asked to select the best summary among those generated by the models, based on criteria such as relevance, fluency, conciseness, informativeness, factual accuracy, and coverage. During the evaluation with ROUGE scores, the 4-bit quantized mBART with LoRA model was found to be effective in generating better Nepali news headlines in comparison to other models and also it was selected 34.05% of the time during the human evaluation, outperforming all other fine-tuned models created for Nepali News headline generation.
Malicious URL classification represents a crucial aspect of cyber security. Although existing work comprises numerous machine learning and deep learning-based URL classification models, most suffer from generalisation and domain-adaptation issues arising from the lack of representative training datasets. Furthermore, these models fail to provide explanations for a given URL classification in natural human language. In this work, we investigate and demonstrate the use of Large Language Models (LLMs) to address this issue. Specifically, we propose an LLM-based one-shot learning framework that uses Chain-of-Thought (CoT) reasoning to predict whether a given URL is benign or phishing. We evaluate our framework using three URL datasets and five state-of-the-art LLMs and show that one-shot LLM prompting indeed provides performances close to supervised models, with GPT 4-Turbo being the best model, followed by Claude 3 Opus. We conduct a quantitative analysis of the LLM explanations and show that most of the explanations provided by LLMs align with the post-hoc explanations of the supervised classifiers, and the explanations have high readability, coherency, and informativeness.
Differential privacy is widely considered the formal privacy for privacy-preserving data analysis due to its robust and rigorous guarantees, with increasingly broad adoption in public services, academia, and industry. Despite originating in the cryptographic context, in this review paper we argue that, fundamentally, differential privacy can be considered a \textit{pure} statistical concept. By leveraging David Blackwell's informativeness theorem, our focus is to demonstrate based on prior work that all definitions of differential privacy can be formally motivated from a hypothesis testing perspective, thereby showing that hypothesis testing is not merely convenient but also the right language for reasoning about differential privacy. This insight leads to the definition of $f$-differential privacy, which extends other differential privacy definitions through a representation theorem. We review techniques that render $f$-differential privacy a unified framework for analyzing privacy bounds in data analysis and machine learning. Applications of this differential privacy definition to private deep learning, private convex optimization, shuffled mechanisms, and U.S.\ Census data are discussed to highlight the benefits of analyzing privacy bounds under this framework compared to existing alternatives.
We establish empirical risk minimization principles for active learning by deriving a family of upper bounds on the generalization error. Aligning with empirical observations, the bounds suggest that superior query algorithms can be obtained by combining both informativeness and representativeness query strategies, where the latter is assessed using integral probability metrics. To facilitate the use of these bounds in application, we systematically link diverse active learning scenarios, characterized by their loss functions and hypothesis classes to their corresponding upper bounds. Our results show that regularization techniques used to constraint the complexity of various hypothesis classes are sufficient conditions to ensure the validity of the bounds. The present work enables principled construction and empirical quality-evaluation of query algorithms in active learning.
Integrating named entity recognition (NER) with automatic speech recognition (ASR) can significantly enhance transcription accuracy and informativeness. In this paper, we introduce WhisperNER, a novel model that allows joint speech transcription and entity recognition. WhisperNER supports open-type NER, enabling recognition of diverse and evolving entities at inference. Building on recent advancements in open NER research, we augment a large synthetic dataset with synthetic speech samples. This allows us to train WhisperNER on a large number of examples with diverse NER tags. During training, the model is prompted with NER labels and optimized to output the transcribed utterance along with the corresponding tagged entities. To evaluate WhisperNER, we generate synthetic speech for commonly used NER benchmarks and annotate existing ASR datasets with open NER tags. Our experiments demonstrate that WhisperNER outperforms natural baselines on both out-of-domain open type NER and supervised finetuning.
Comparing datasets is a fundamental task in machine learning, essential for various learning paradigms; from evaluating train and test datasets for model generalization to using dataset similarity for detecting data drift. While traditional notions of dataset distances offer principled measures of similarity, their utility has largely been assessed through prediction error minimization. However, in Predict-then-Optimize (PtO) frameworks, where predictions serve as inputs for downstream optimization tasks, model performance is measured through decision regret minimization rather than prediction error minimization. In this work, we (i) show that traditional dataset distances, which rely solely on feature and label dimensions, lack informativeness in the PtO context, and (ii) propose a new dataset distance that incorporates the impacts of downstream decisions. Our results show that this decision-aware dataset distance effectively captures adaptation success in PtO contexts, providing a PtO adaptation bound in terms of dataset distance. Empirically, we show that our proposed distance measure accurately predicts transferability across three different PtO tasks from the literature.