Large Language Models (LLMs) play a central role in modern artificial intelligence, yet their development has been primarily focused on English, resulting in limited support for other languages. We present PLLuM (Polish Large Language Model), the largest open-source family of foundation models tailored specifically for the Polish language. Developed by a consortium of major Polish research institutions, PLLuM addresses the need for high-quality, transparent, and culturally relevant language models beyond the English-centric commercial landscape. We describe the development process, including the construction of a new 140-billion-token Polish text corpus for pre-training, a 77k custom instructions dataset, and a 100k preference optimization dataset. A key component is a Responsible AI framework that incorporates strict data governance and a hybrid module for output correction and safety filtering. We detail the models' architecture, training procedures, and alignment techniques for both base and instruction-tuned variants, and demonstrate their utility in a downstream task within public administration. By releasing these models publicly, PLLuM aims to foster open research and strengthen sovereign AI technologies in Poland.

The critical shortage of mental health services due to workforce limitations and logistical barriers, especially in underserved areas designated by the Health Resources & Services Administration (HRSA) 1, highlights the urgent need for accessible and scalable solutions. Traditional services often fail to address the diverse needs of individuals experiencing anxiety, prompting many, especially younger populations, to seek alternative emotional and psychological support online. While digital platforms offer immediate access, unregulated online interactions, including those with generative AI, may disseminate misleading information or inappropriate advice, potentially exacerbating anxiety symptoms (Tobias & Ito, 2021). Despite the great potential of generative AI to supplement mental health services, its deployment poses potentially significant risks. Unlike clinical practitioners, LLMs are not inherently equipped to manage emotionally complex or vulnerable conversations, which are critical to therapeutic relationships that create positive clinical outcomes (Rogers, 1957; Wampold, 2015).

Effective communication about breast and cervical cancers remains a persistent health challenge, with significant gaps in public understanding of cancer prevention, screening, and treatment, potentially leading to delayed diagnoses and inadequate treatments. This study evaluates the capabilities and limitations of Large Language Models (LLMs) in generating accurate, safe, and accessible cancer-related information to support patient understanding. We evaluated five general-purpose and three medical LLMs using a mixed-methods evaluation framework across linguistic quality, safety and trustworthiness, and communication accessibility and affectiveness. Our approach utilized quantitative metrics, qualitative expert ratings, and statistical analysis using Welch's ANOVA, Games-Howell, and Hedges' g. Our results show that general-purpose LLMs produced outputs of higher linguistic quality and affectiveness, while medical LLMs demonstrate greater communication accessibility. However, medical LLMs tend to exhibit higher levels of potential harm, toxicity, and bias, reducing their performance in safety and trustworthiness. Our findings indicate a duality between domain-specific knowledge and safety in health communications. The results highlight the need for intentional model design with targeted improvements, particularly in mitigating harm and bias, and improving safety and affectiveness. This study provides a comprehensive evaluation of LLMs for cancer communication, offering critical insights for improving AI-generated health content and informing future development of accurate, safe, and accessible digital health tools.

Large Language Models (LLMs) have significantly advanced natural language processing (NLP) tasks but also pose ethical and societal risks due to their propensity to generate harmful content. To address this, various approaches have been developed to safeguard LLMs from producing unsafe content. However, existing methods have limitations, including the need for training specific control models and proactive intervention during text generation, that lead to quality degradation and increased computational overhead. To mitigate those limitations, we propose LLMSafeGuard, a lightweight framework to safeguard LLM text generation in real-time. LLMSafeGuard integrates an external validator into the beam search algorithm during decoding, rejecting candidates that violate safety constraints while allowing valid ones to proceed. We introduce a similarity based validation approach, simplifying constraint introduction and eliminating the need for control model training. Additionally, LLMSafeGuard employs a context-wise timing selection strategy, intervening LLMs only when necessary. We evaluate LLMSafeGuard on two tasks, detoxification and copyright safeguarding, and demonstrate its superior performance over SOTA baselines. For instance, LLMSafeGuard reduces the average toxic score of. LLM output by 29.7% compared to the best baseline meanwhile preserving similar linguistic quality as natural output in detoxification task. Similarly, in the copyright task, LLMSafeGuard decreases the Longest Common Subsequence (LCS) by 56.2% compared to baselines. Moreover, our context-wise timing selection strategy reduces inference time by at least 24% meanwhile maintaining comparable effectiveness as validating each time step. LLMSafeGuard also offers tunable parameters to balance its effectiveness and efficiency.

We propose CHRT (Control Hidden Representation Transformation) - a controlled language generation framework that steers large language models to generate text pertaining to certain attributes (such as toxicity). CHRT gains attribute control by modifying the hidden representation of the base model through learned transformations. We employ a contrastive-learning framework to learn these transformations that can be combined to gain multi-attribute control. The effectiveness of CHRT is experimentally shown by comparing it with seven baselines over three attributes. CHRT outperforms all the baselines in the task of detoxification, positive sentiment steering, and text simplification while minimizing the loss in linguistic qualities. Further, our approach has the lowest inference latency of only 0.01 seconds more than the base model, making it the most suitable for high-performance production environments. We open-source our code and release two novel datasets to further propel controlled language generation research.

Recent advances in generative models for language have enabled the creation of convincing synthetic text or deepfake text. Prior work has demonstrated the potential for misuse of deepfake text to mislead content consumers. Therefore, deepfake text detection, the task of discriminating between human and machine-generated text, is becoming increasingly critical. Several defenses have been proposed for deepfake text detection. However, we lack a thorough understanding of their real-world applicability. In this paper, we collect deepfake text from 4 online services powered by Transformer-based tools to evaluate the generalization ability of the defenses on content in the wild. We develop several low-cost adversarial attacks, and investigate the robustness of existing defenses against an adaptive attacker. We find that many defenses show significant degradation in performance under our evaluation scenarios compared to their original claimed performance. Our evaluation shows that tapping into the semantic information in the text content is a promising approach for improving the robustness and generalization performance of deepfake text detection schemes.

Poetry generation has long been a challenge for artificial intelligence. In the scope of Japanese poetry generation, many researchers have paid attention to Haiku generation, but few have focused on Waka generation. To further explore the creative potential of natural language generation systems in Japanese poetry creation, we propose a novel Waka generation model, WakaVT, which automatically produces Waka poems given user-specified keywords. Firstly, an additive mask-based approach is presented to satisfy the form constraint. Secondly, the structures of Transformer and variational autoencoder are integrated to enhance the quality of generated content. Specifically, to obtain novelty and diversity, WakaVT employs a sequence of latent variables, which effectively captures word-level variability in Waka data. To improve linguistic quality in terms of fluency, coherence, and meaningfulness, we further propose the fused multilevel self-attention mechanism, which properly models the hierarchical linguistic structure of Waka. To the best of our knowledge, we are the first to investigate Waka generation with models based on Transformer and/or variational autoencoder. Both objective and subjective evaluation results demonstrate that our model outperforms baselines significantly.

Distributed word representations have a rising interest in NLP community. Most of existing models assume only one vector for each individual word, which ignores polysemy and thus degrades their effectiveness for downstream tasks. To address this problem, some recent work adopts multi-prototype models to learn multiple embeddings per word type. In this paper, we distinguish the different senses of each word by their latent topics. We present a general architecture to learn the word and topic embeddings efficiently, which is an extension to the Skip-Gram model and can model the interaction between words and topics simultaneously. The experiments on the word similarity and text classification tasks show our model outperforms state-of-the-art methods.

Distributed word representations have a rising interest in NLP community. Most of existing models assume only one vector for each individual word, which ignores polysemy and thus degrades their effectiveness for downstream tasks. To address this problem, some recent work adopts multi-prototype models to learn multiple embeddings per word type. In this paper, we distinguish the different senses of each word by their latent topics. We present a general architecture to learn the word and topic embeddings efficiently, which is an extension to the Skip-Gram model and can model the interaction between words and topics simultaneously. The experiments on the word similarity and text classification tasks show our model outperforms state-of-the-art methods.

Distributed word representations have a rising interest in NLP community. Most of existing models assume only one vector for each individual word, which ignores polysemy and thus degrades their effectiveness for downstream tasks. To address this problem, some recent work adopts multi-prototype models to learn multiple embeddings per word type. In this paper, we distinguish the different senses of each word by their latent topics. We present a general architecture to learn the word and topic embeddings efficiently, which is an extension to the Skip-Gram model and can model the interaction between words and topics simultaneously. The experiments on the word similarity and text classification tasks show our model outperforms state-of-the-art methods.