ELOQUENT is a set of shared tasks that aims to create easily testable high-level criteria for evaluating generative language models. Sensemaking is one such shared task. In Sensemaking, we try to assess how well generative models ``make sense out of a given text'' in three steps inspired by exams in a classroom setting: (1) Teacher systems should prepare a set of questions, (2) Student systems should answer these questions, and (3) Evaluator systems should score these answers, all adhering rather strictly to a given set of input materials. We report on the 2025 edition of Sensemaking, where we had 7 sources of test materials (fact-checking analyses of statements, textbooks, transcribed recordings of a lecture, and educational videos) spanning English, German, Ukrainian, and Czech languages. This year, 4 teams participated, providing us with 2 Teacher submissions, 2 Student submissions, and 2 Evaluator submissions. We added baselines for Teacher and Student using commercial large language model systems. We devised a fully automatic evaluation procedure, which we compare to a minimalistic manual evaluation. We were able to make some interesting observations. For the first task, the creation of questions, better evaluation strategies will still have to be devised because it is difficult to discern the quality of the various candidate question sets. In the second task, question answering, the LLMs examined overall perform acceptably, but restricting their answers to the given input texts remains problematic. In the third task, evaluation of question answers, our adversarial tests reveal that systems using the LLM-as-a-Judge paradigm erroneously rate both garbled question-answer pairs and answers to mixed-up questions as acceptable.

Text data augmentation is a widely used strategy for mitigating data sparsity in natural language processing (NLP), particularly in low-resource settings where limited samples hinder effective semantic modeling. While augmentation can improve input diversity and downstream interpretability, existing techniques often lack mechanisms to ensure semantic preservation during large -scale or iterative generation, leadin g to redundancy and instability. This work introduces a principled evaluation framework for large language model (LLM) based text augmentation, comprising two components: (1) Scalability Analysis, which measures semantic consistency as augmentation volume increases, and (2) Iterative Augmentation with Summarization Refinement (IASR), which evaluates semantic drift across recursive paraphrasing cycles. Empirical evaluations across state -of-the -art LLMs show that GPT-3.5 Turbo achieve d the best balance of semantic fidelity, diversity, and generation efficiency. Applied to a real -world topic modeling task using BERTopic with GPT-enhanced few -shot labeling, the proposed approach results in a 400% increase in topic granularity and complete elimination of topic overlaps. These findings validate d the utility of the proposed frameworks for structured evaluation of LLM -based augmentation in practical NLP pipelines.

The SaaS paradigm has revolutionized software distribution by offering flexible pricing options to meet diverse customer needs. However, the rapid expansion of the SaaS market has introduced significant complexity for DevOps teams, who must manually manage and evolve pricing structures, an approach that is both time-consuming and prone to errors. The absence of automated tools for pricing analysis restricts the ability to efficiently evaluate, optimize, and scale these models. This paper proposes leveraging intelligent pricing (iPricing), dynamic, machine-readable pricing models, as a solution to these challenges. Intelligent pricing enables competitive analysis, streamlines operational decision-making, and supports continuous pricing evolution in response to market dynamics, leading to improved efficiency and accuracy. We present an LLM-driven approach that automates the transformation of static HTML pricing into iPricing, significantly improving efficiency and consistency while minimizing human error. Our implementation, AI4Pricing2Yaml, features a basic Information Extractor that uses web scraping and LLMs technologies to extract essential pricing components, plans, features, usage limits, and add-ons, from SaaS websites. Validation against a dataset of 30 distinct commercial SaaS, encompassing over 150 intelligent pricings, demonstrates the system's effectiveness in extracting the desired elements across all steps. However, challenges remain in addressing hallucinations, complex structures, and dynamic content. This work highlights the potential of automating intelligent pricing transformation to streamline SaaS pricing management, offering implications for improved consistency and scalability in an increasingly intricate pricing landscape. Future research will focus on refining extraction capabilities and enhancing the system's adaptability to a wider range of SaaS websites.

In many applications, decision-makers must choose between multiple predictive models that may all be miscalibrated. Which model (i.e., predictor) is more "useful" in downstream decision tasks? To answer this, our first contribution introduces the notion of the informativeness gap between any two predictors, defined as the maximum normalized payoff advantage one predictor offers over the other across all decision-making tasks. Our framework strictly generalizes several existing notions: it subsumes U-Calibration [KLST-23] and Calibration Decision Loss [HW-24], which compare a miscalibrated predictor to its calibrated counterpart, and it recovers Blackwell informativeness [Bla-51, Bla-53] as a special case when both predictors are perfectly calibrated. Our second contribution is a dual characterization of the informativeness gap, which gives rise to a natural informativeness measure that can be viewed as a relaxed variant of the earth mover's distance (EMD) between two prediction distributions. We show that this measure satisfies natural desiderata: it is complete and sound, and it can be estimated sample-efficiently in the prediction-only access setting. Along the way, we also obtain novel combinatorial structural results when applying this measure to perfectly calibrated predictors.

This paper presents an intervention study on the effects of the combined methods of (1) the Socratic method, (2) Chain of Thought (CoT) reasoning, (3) simplified gamification and (4) formative feedback on university students' Maths learning driven by large language models (LLMs). We call our approach Mathematics Explanations through Games by AI LLMs (MEGA). Some students struggle with Maths and as a result avoid Math-related discipline or subjects despite the importance of Maths across many fields, including signal processing. Oftentimes, students' Maths difficulties stem from suboptimal pedagogy. We compared the MEGA method to the traditional step-by-step (CoT) method to ascertain which is better by using a within-group design after randomly assigning questions for the participants, who are university students. Samples (n=60) were randomly drawn from each of the two test sets of the Grade School Math 8K (GSM8K) and Mathematics Aptitude Test of Heuristics (MATH) datasets, based on the error margin of 11%, the confidence level of 90%, and a manageable number of samples for the student evaluators. These samples were used to evaluate two capable LLMs at length (Generative Pretrained Transformer 4o (GPT4o) and Claude 3.5 Sonnet) out of the initial six that were tested for capability. The results showed that students agree in more instances that the MEGA method is experienced as better for learning for both datasets. It is even much better than the CoT (47.5% compared to 26.67%) in the more difficult MATH dataset, indicating that MEGA is better at explaining difficult Maths problems.

Coreference Resolution systems are typically evaluated on benchmarks containing small- to medium-scale documents. When it comes to evaluating long texts, however, existing benchmarks, such as LitBank, remain limited in length and do not adequately assess system capabilities at the book scale, i.e., when co-referring mentions span hundreds of thousands of tokens. To fill this gap, we first put forward a novel automatic pipeline that produces high-quality Coreference Resolution annotations on full narrative texts. Then, we adopt this pipeline to create the first book-scale coreference benchmark, BOOKCOREF, with an average document length of more than 200,000 tokens. We carry out a series of experiments showing the robustness of our automatic procedure and demonstrating the value of our resource, which enables current long-document coreference systems to gain up to +20 CoNLL-F1 points when evaluated on full books. Moreover, we report on the new challenges introduced by this unprecedented book-scale setting, highlighting that current models fail to deliver the same performance they achieve on smaller documents. We release our data and code to encourage research and development of new book-scale Coreference Resolution systems at https://github.com/sapienzanlp/bookcoref.

This submission to the binary AI detection task is based on a modular stylometric pipeline, where: public spaCy models are used for text preprocessing (including tokenisation, named entity recognition, dependency parsing, part-of-speech tagging, and morphology annotation) and extracting several thousand features (frequencies of n-grams of the above linguistic annotations); light-gradient boosting machines are used as the classifier. We collect a large corpus of more than 500 000 machine-generated texts for the classifier's training. We explore several parameter options to increase the classifier's capacity and take advantage of that training set. Our approach follows the non-neural, computationally inexpensive but explainable approach found effective previously.

Human feedback is increasingly used across diverse applications like training AI models, developing recommender systems, and measuring public opinion -- with granular feedback often being preferred over binary feedback for its greater informativeness. While it is easy to accurately estimate a population's distribution of feedback given feedback from a large number of individuals, cost constraints typically necessitate using smaller groups. A simple method to approximate the population distribution is regularized averaging: compute the empirical distribution and regularize it toward a prior. Can we do better? As we will discuss, the answer to this question depends on feedback granularity. Suppose one wants to predict a population's distribution of feedback using feedback from a limited number of individuals. We show that, as feedback granularity increases, one can substantially improve upon predictions of regularized averaging by combining individuals' feedback in ways more sophisticated than regularized averaging. Our empirical analysis using questions on social attitudes confirms this pattern. In particular, with binary feedback, sophistication barely reduces the number of individuals required to attain a fixed level of performance. By contrast, with five-point feedback, sophisticated methods match the performance of regularized averaging with about half as many individuals.

This thesis addresses challenges related to data and parameter efficiency in neural language models, with a focus on representation analysis and the introduction of new optimization techniques. The first part examines the properties and dynamics of language representations within neural models, emphasizing their significance in enhancing robustness and generalization. It proposes innovative approaches based on representation smoothness, including regularization strategies that utilize Jacobian and Hessian matrices to stabilize training and mitigate sensitivity to input perturbations. The second part focuses on methods to significantly enhance data and parameter efficiency by integrating active learning strategies with parameter-efficient fine-tuning, guided by insights from representation smoothness analysis. It presents smoothness-informed early-stopping techniques designed to eliminate the need for labeled validation sets and proposes innovative combinations of active learning and parameter-efficient fine-tuning to reduce labeling efforts and computational resources. Extensive experimental evaluations across various NLP tasks demonstrate that these combined approaches substantially outperform traditional methods in terms of performance, stability, and efficiency. The third part explores weak supervision techniques enhanced by in-context learning to effectively utilize unlabeled data, further reducing dependence on extensive labeling. It shows that using in-context learning as a mechanism for weak supervision enables models to better generalize from limited labeled data by leveraging unlabeled examples more effectively during training. Comprehensive empirical evaluations confirm significant gains in model accuracy, adaptability, and robustness, especially in low-resource settings and dynamic data environments.

Generative Artificial Intelligence (GAI) has rapidly emerged as a transformative force in aquaculture, enabling intelligent synthesis of multimodal data, including text, images, audio, and simulation outputs for smarter, more adaptive decision-making. As the aquaculture industry shifts toward data-driven, automation and digital integration operations under the Aquaculture 4.0 paradigm, GAI models offer novel opportunities across environmental monitoring, robotics, disease diagnostics, infrastructure planning, reporting, and market analysis. This review presents the first comprehensive synthesis of GAI applications in aquaculture, encompassing foundational architectures (e.g., diffusion models, transformers, and retrieval augmented generation), experimental systems, pilot deployments, and real-world use cases. We highlight GAI's growing role in enabling underwater perception, digital twin modeling, and autonomous planning for remotely operated vehicle (ROV) missions. We also provide an updated application taxonomy that spans sensing, control, optimization, communication, and regulatory compliance. Beyond technical capabilities, we analyze key limitations, including limited data availability, real-time performance constraints, trust and explainability, environmental costs, and regulatory uncertainty. This review positions GAI not merely as a tool but as a critical enabler of smart, resilient, and environmentally aligned aquaculture systems.