Electric power marketing telephone customer service primarily communicates with customers via phone calls to understand their electricity usage needs, provide consultations, process service applications, and handle complaints [1]. Ensuring timely and effective responses is essential throughout the service process. However, current systems (e.g., 95598, the customer service hotline of State Grid Corporation of China) often suffer from poor user experience, delayed responses, and inaccurate information[2] [3]. These traditional systems rely heavily on fixed procedures and templates, lacking the flexibility to address complex and diverse customer demands. This limitation is particularly pronounced in the highly specialized field of electric power marketing, where slow response times and insufficiently tailored solutions negatively impact service quality. Although human agents can complement these systems by managing more complex issues, they also face significant challenges, such as high workloads during peak periods, delayed response times, and inconsistent levels of professional knowledge and expertise. As a result, it is difficult to guarantee consistent and high-quality service for all customers.

Online property booking platforms are widely used and rely heavily on consistent, up-to-date information about accommodation facilities, often sourced from third-party providers. However, these external data sources are frequently affected by incomplete or inconsistent details, which can frustrate users and result in a loss of market. In response to these challenges, we present an industrial case study involving the integration of Large Language Models (LLMs) into CALEIDOHOTELS, a property reservation platform developed by FERVENTO. We evaluate two well-known LLMs in this context: Mistral 7B, fine-tuned with QLoRA, and Mixtral 8x7B, utilized with a refined system prompt. Both models were assessed based on their ability to generate consistent and homogeneous descriptions while minimizing hallucinations. Mixtral 8x7B outperformed Mistral 7B in terms of completeness (99.6% vs. 93%), precision (98.8% vs. 96%), and hallucination rate (1.2% vs. 4%), producing shorter yet more concise content (249 vs. 277 words on average). However, this came at a significantly higher computational cost: 50GB VRAM and $1.61/hour versus 5GB and $0.16/hour for Mistral 7B. Our findings provide practical insights into the trade-offs between model quality and resource efficiency, offering guidance for deploying LLMs in production environments and demonstrating their effectiveness in enhancing the consistency and reliability of accommodation data.

The CSS was accompanied by a natural language explanation of the scores. The LLM judge role used GPT-4.0 by OpenAI. Evaluation by Human Experts Each encounter pair in which the top diagnosis of AI and clinician did not match was evaluated by a board-certified physician with access to medical reference material. Blinding the physician to the origin of the documentation proved impractical, as the AI-based notes were highly consistent and thus easily recognized within a few pairs. The physician was asked to determine the cause of the disagreement between the documents, whether AI or the physician was more likely to be correct, whether it was not possible to determine which diagnosis was more appropriate, and whether the diagnoses did, in fact, match. Similarity and Style Metrics To evaluate how similar-or different the AI-generated (Doctronic) and clinician-generated SOAP notes were, we followed a two-step process. First, we assessed surface-level textual similarity using three standard statistical metrics: (1) TF IDF cosine similarity, which transforms each note into a weighted term-frequency vector and measures the cosine of the angle between them to capture word-frequency alignment; (2) the Jaccard index, which is the ratio of the intersection to the union of lowercased token sets, ranging from 0 (no overlap) to 1 (identical token sets); and (3) the Levenshtein ratio, a normalized edit-distance score based on character-level insertions, deletions, and substitutions that quantifies textual similarity on a 0-1 scale. These analyses demonstrated only minimal alignment in phrasing, formatting, and vocabulary. Then, to probe contextual and semantic similarity, we generated embeddings for each note using OpenAI's text embedding 3 small model and two versions of Biobert,

Conversational recommender systems (CRS) enhance user experience through multi-turn interactions, yet evaluating CRS remains challenging. User simulators can provide comprehensive evaluations through interactions with CRS, but building realistic and diverse simulators is difficult. While recent work leverages large language models (LLMs) to simulate user interactions, they still fall short in emulating individual real users across diverse scenarios and lack explicit rating mechanisms for quantitative evaluation. To address these gaps, we propose RecUserSim, an LLM agent-based user simulator with enhanced simulation realism and diversity while providing explicit scores. RecUserSim features several key modules: a profile module for defining realistic and diverse user personas, a memory module for tracking interaction history and discovering unknown preferences, and a core action module inspired by Bounded Rationality theory that enables nuanced decision-making while generating more fine-grained actions and personalized responses. To further enhance output control, a refinement module is designed to fine-tune final responses. Experiments demonstrate that RecUserSim generates diverse, controllable outputs and produces realistic, high-quality dialogues, even with smaller base LLMs. The ratings generated by RecUserSim show high consistency across different base LLMs, highlighting its effectiveness for CRS evaluation.

It is crucial that robots' performance can be improved after deployment, as they are inherently likely to encounter novel scenarios never seen before. This paper presents an innovative solution: an interactive learning-based robot system powered by a Multi-modal Large Language Model(MLLM). A key feature of our system is its ability to learn from natural dialogues with non-expert users. We also propose chain of question to clarify the exact intent of the question before providing an answer and dual-modality retrieval modules to leverage these interaction events to avoid repeating same mistakes, ensuring a seamless user experience before model updates, which is in contrast to current mainstream MLLM-based robotic systems. Our system marks a novel approach in robotics by integrating interactive learning, paving the way for superior adaptability and performance in diverse environments. We demonstrate the effectiveness and improvement of our method through experiments, both quantitively and qualitatively.

--Conventional augmentative and alternative communication (AAC) systems and language-learning platforms often fail to adapt in real time to the user's cognitive and linguistic needs, especially in neurological conditions such as post-stroke aphasia or amyotrophic lateral sclerosis. Recent advances in noninvasive electroencephalography (EEG)-based brain-computer interfaces (BCIs) and transformer-based large language models (LLMs) offer complementary strengths: BCIs capture users' neural intent with low fatigue, while LLMs generate contextually tailored language content. Objective: We propose and evaluate a novel hybrid framework that leverages real-time EEG signals to drive an LLM-powered language rehabilitation assistant. This system aims to: (1) enable users with severe speech or motor impairments to navigate language-learning modules via mental commands; (2) dynamically personalize vocabulary, sentence-construction exercises, and corrective feedback; and (3) monitor neural markers of cognitive effort to adjust task difficulty on the fly. All individuals have the right to self-expression, social participation, and the agency to impact their environment. For individuals with complex communication needs, augmentative and alternative communication (AAC) systems provide critical tools to facilitate communication. However, traditional AAC methods--such as printed communication boards or eye gaze devices--may not be accessible for individuals with severe speech and physical impairments (SSPI).

With the amount and complexity of information produced increasing at staggering rates, information visualization is being utilized to enable people to understand and analyze information. Over the years, many techniques have been developed for creating information visualizations of different types of data. Information visualization can be created using various tools (for example, Tableau [1]), libraries in many programming languages (for example, matplotlib [2]), as well as scripts (for example, Vega-lite [3]). However, the complexity of these tools, libraries, and scripts can pose a barrier, especially for people without a strong background in data science or programming. To address this, automation of visualization creation using artificial intelligence techniques has also been explored [4]. Natural language interfaces allow users to generate visualizations using simple and intuitive commands. The integration of natural language processing into data visualization tools significantly improves the efficiency of data analysis. Analysts can now focus more on interpreting the data rather than the technicalities of creating visualizations.

Reinforcement learning has proven its effectiveness in enhancing the reasoning capabilities of large language models. Recent research efforts have progressively extended this paradigm to multimodal reasoning tasks. Due to the inherent complexity and diversity of multimodal tasks, especially in semantic content and problem formulations, existing models often exhibit unstable performance across various domains and difficulty levels. To address these limitations, we propose VL-Cogito, an advanced multimodal reasoning model trained via a novel multi-stage Progressive Curriculum Reinforcement Learning (PCuRL) framework. PCuRL systematically guides the model through tasks of gradually increasing difficulty, substantially improving its reasoning abilities across diverse multimodal contexts. The framework introduces two key innovations: (1) an online difficulty soft weighting mechanism, dynamically adjusting training difficulty across successive RL training stages; and (2) a dynamic length reward mechanism, which encourages the model to adaptively regulate its reasoning path length according to task complexity, thus balancing reasoning efficiency with correctness. Experimental evaluations demonstrate that VL-Cogito consistently matches or surpasses existing reasoning-oriented models across mainstream multimodal benchmarks spanning mathematics, science, logic, and general understanding, validating the effectiveness of our approach.

Large Reasoning Models (LRMs) have achieved significant advances in mathematical reasoning and formal logic tasks. However, their tendency to generate lengthy chain-of-thought sequences leads to substantial memory overhead during inference. We observe that LRMs frequently produce highly similar intermediate reasoning steps, which correspond to similar KV cache states across layers. Motivated by this observation, we propose MemShare, a novel KV cache management approach that effectively reduces memory overhead. MemShare employs a collaborative filtering algorithm to efficiently identify reusable KV cache blocks and enables zero copy cache reuse to significantly reduce memory overhead, improve throughput while maintaining accuracy. Experimental results demonstrate that MemShare delivers up to 84.79\% improvement in throughput while maintaining better accuracy compared to existing KV cache management methods.

Human action recognition plays a critical role in healthcare and medicine, supporting applications such as patient behavior monitoring, fall detection, surgical robot supervision, and procedural skill assessment. While traditional models like CNNs and RNNs have achieved moderate success, they often struggle to generalize across diverse and complex actions. Recent advancements in vision-language models, especially the transformer-based CLIP model, offer promising capabilities for generalizing action recognition from video data. In this work, we evaluate CLIP on the UCF-101 dataset and systematically analyze its performance under three masking strategies: (1) percentage-based and shape-based black masking at 10%, 30%, and 50%, (2) feature-specific masking to suppress bias-inducing elements, and (3) isolation masking that retains only class-specific regions. Our results reveal that CLIP exhibits inconsistent behavior and frequent misclassifications, particularly when essential visual cues are obscured. To overcome these limitations, we propose incorporating class-specific noise, learned via a custom loss function, to reinforce attention to class-defining features. This enhancement improves classification accuracy and model confidence while reducing bias. We conclude with a discussion on the challenges of applying such models in clinical domains and outline directions for future work to improve generalizability across domain-independent healthcare scenarios.