Artificial Intelligence models have demonstrated significant success in diagnosing skin diseases, including cancer, showing the potential to assist clinicians in their analysis. However, the interpretability of model predictions must be significantly improved before they can be used in practice. To this end, we explore the combination of two promising approaches: Multimodal Large Language Models (MLLMs) and quantitative attribute usage. MLLMs offer a potential avenue for increased interpretability, providing reasoning for diagnosis in natural language through an interactive format. Separately, a number of quantitative attributes that are related to lesion appearance (e.g., lesion area) have recently been found predictive of malignancy with high accuracy. Predictions grounded as a function of such concepts have the potential for improved interpretability. We provide evidence that MLLM embedding spaces can be grounded in such attributes, through fine-tuning to predict their values from images. Concretely, we evaluate this grounding in the embedding space through an attribute-specific content-based image retrieval case study using the SLICE-3D dataset.

Multimodal Large Language Models (MLLMs) emerge as a unified interface to address a multitude of tasks, ranging from NLP to computer vision. Despite showcasing state-of-the-art results in many benchmarks, a long-standing issue is the tendency of MLLMs to hallucinate, that is to generate answers to the user's query that are not reflected in the visual input. In this paper, we address the problem of hallucinations as an alignment problem, seeking to steer the MLLM so that it prefers generating content without hallucinations. In contrast to recent approaches that require complicated pipelines to build synthetic preference data for alignment training, often relying on proprietary models, we capitalize on the well-known CHAIR metric, originally proposed to gauge the degree of hallucinations in image captioning. Given a pair of generated answers, we leverage CHAIR to distinguish winner and loser options (i.e., non-hallucinated and hallucinated samples) and fine-tune off-the-shelf MLLMs via Direct Preference Optimization (DPO). The resulting method, which we refer to as CHAIR-DPO, effectively diminishes the amount of hallucinated answers on several hallucination benchmarks, demonstrating the effectiveness of fine-tuning the MLLM with a CHAIR-based reward. Source code and trained models are publicly available at https://github.com/aimagelab/CHAIR-DPO.

The rapid advancement of large language models (LLMs) has driven their adoption across diverse domains, yet their ability to generate harmful content poses significant safety challenges. While extensive research has focused on mitigating harmful outputs, such efforts often come at the cost of excessively rejecting harmless prompts. Striking a balance among safety, over-refusal, and utility remains a critical challenge. In this work, we introduce IntentionReasoner, a novel safeguard mechanism that leverages a dedicated guard model to perform intent reasoning, multi-level safety classification, and query rewriting to neutralize potentially harmful intent in edge-case queries. Specifically, we first construct a comprehensive dataset comprising approximately 163,000 queries, each annotated with intent reasoning, safety labels, and rewritten versions. Supervised fine-tuning is then applied to equip the guard model with foundational capabilities in format adherence, intent analysis, and safe rewriting. Finally, we apply a tailored multi-reward optimization strategy that integrates rule-based heuristics and reward model signals within a reinforcement learning framework to further enhance performance. Extensive experiments show that IntentionReasoner excels in multiple safeguard benchmarks, generation quality evaluations, and jailbreak attack scenarios, significantly enhancing safety while effectively reducing over-refusal rates and improving the quality of responses.

Designing crystal materials with desired physicochemical properties remains a fundamental challenge in materials science. While large language models (LLMs) have demonstrated strong in-context learning (ICL) capabilities, existing LLM-based crystal generation approaches are limited to zero-shot scenarios and are unable to benefit from few-shot scenarios. In contrast, human experts typically design new materials by modifying relevant known structures which aligns closely with the few-shot ICL paradigm. Motivated by this, we propose CrystalICL, a novel model designed for few-shot crystal generation. Specifically, we introduce a space-group based crystal tokenization method, which effectively reduces the complexity of modeling crystal symmetry in LLMs. We further introduce a condition-structure aware hybrid instruction tuning framework and a multi-task instruction tuning strategy, enabling the model to better exploit ICL by capturing structure-property relationships from limited data. Extensive experiments on four crystal generation benchmarks demonstrate the superiority of CrystalICL over the leading baseline methods on conditional and unconditional generation tasks.

Medical image segmentation has advanced rapidly over the past two decades, largely driven by deep learning, which has enabled accurate and efficient delineation of cells, tissues, organs, and pathologies across diverse imaging modalities. This progress raises a fundamental question: to what extent have current models overcome persistent challenges, and what gaps remain? In this work, we provide an in-depth review of medical image segmentation, tracing its progress and key developments over the past decade. We examine core principles, including multiscale analysis, attention mechanisms, and the integration of prior knowledge, across the encoder, bottleneck, skip connections, and decoder components of segmentation networks. Our discussion is organized around seven key dimensions: (1) the shift from supervised to semi-/unsupervised learning, (2) the transition from organ segmentation to lesion-focused tasks, (3) advances in multi-modality integration and domain adaptation, (4) the role of foundation models and transfer learning, (5) the move from deterministic to probabilistic segmentation, (6) the progression from 2D to 3D and 4D segmentation, and (7) the trend from model invocation to segmentation agents. Together, these perspectives provide a holistic overview of the trajectory of deep learning-based medical image segmentation and aim to inspire future innovation. To support ongoing research, we maintain a continually updated repository of relevant literature and open-source resources at https://github.com/apple1986/medicalSegReview

--In this case study, we present a data-efficient point cloud segmentation pipeline and training framework for robust segmentation of unimproved roads and seven other classes. Our method employs a two-stage training framework: first, a projection-based convolutional neural network is pre-trained on a mixture of public urban datasets and a small, curated in-domain dataset; then, a lightweight prediction head is fine-tuned exclusively on in-domain data. Along the way, we explore the application of Point Prompt Training to batch normalization layers and the effects of Manifold Mixup as a regularizer within our pipeline. We also explore the effects of incorporating histogram-normalized ambients to further boost performance. Using only 50 labeled point clouds from our target domain, we show that our proposed training approach improves mean Intersection-over-Union from 33.5% to 51.8% and the overall accuracy from 85.5% to 90.8%, when compared to naive training on the in-domain data. Crucially, our results demonstrate that pre-training across multiple datasets is key to improving generalization and enabling robust segmentation under limited in-domain supervision. Overall, this study demonstrates a practical framework for robust 3D semantic segmentation in challenging, low-data scenarios. Semantic segmentation of 3D point clouds is a foundational task for scene understanding, enabling a range of downstream applications such as autonomous route planning and infrastructure inspection. Despite significant progress in this field, most state-of-the-art segmentation models rely heavily on the availability of large, labeled training datasets. However, generating labeled point cloud data remains a substantial bottleneck: manual annotation is both labor-intensive and time-consuming, requiring over 30 minutes per scan on average in our experiments. This challenge makes it impractical to recreate large-scale datasets, commonly containing over 25,000 scans, for new or underrepresented environments.

The CRISPR-Cas genome editing system has rapidly become an indispensable tool across biotechnology and medicine, enabling targeted DNA modifications with unprecedented ease. A single-guide RNA (sgRNA, or simply gRNA) directs the Cas nuclease (such as Cas9 or Cas12a) to a complementary genomic sequence, where the nuclease induces a double-strand break or nucleotide modification. The efficiency and specificity of this process are largely dictated by the gRNA sequence and its interactions with both the target DNA and the cellular environment. Designing optimal gRNAs is therefore critical for successful editing outcomes. Early gRNA design relied on empirical rules and modest machine learning models, but these approaches often struggled to capture the complex determinants of gRNA activity and off-target effects. In recent years, artificial intelligence (AI) - particularly deep learning - has been leveraged to overcome these limitations, learning predictive features from large-scale CRISPR datasets and outperforming previous rule-based methods in guide efficacy prediction[1, 2]. Deep learning models can ingest not only the gRNA and target DNA sequences but also additional contextual information (e.g.

Purpose: Spiking neural networks (SNNs) have recently gained attention as energy-efficient, biologically plausible alternatives to conventional deep learning models. Their application in high-stakes biomedical imaging remains almost entirely unexplored. Methods: This study introduces SNNDeep, the first tailored SNN specifically optimized for binary classification of liver health status from computed tomography (CT) features. To ensure clinical relevance and broad generalizability, the model was developed and evaluated using the Task03\Liver dataset from the Medical Segmentation Decathlon (MSD), a standardized benchmark widely used for assessing performance across diverse medical imaging tasks. We benchmark three fundamentally different learning algorithms, namely Surrogate Gradient Learning, the Tempotron rule, and Bio-Inspired Active Learning across three architectural variants: a fully customized low-level model built from scratch, and two implementations using leading SNN frameworks, i.e., snnTorch and SpikingJelly. Hyperparameter optimization was performed using Optuna. Results: Our results demonstrate that the custom-built SNNDeep consistently outperforms framework-based implementations, achieving a maximum validation accuracy of 98.35%, superior adaptability across learning rules, and significantly reduced training overhead. Conclusion:This study provides the first empirical evidence that low-level, highly tunable SNNs can surpass standard frameworks in medical imaging, especially in data-limited, temporally constrained diagnostic settings, thereby opening a new pathway for neuro-inspired AI in precision medicine.

While code large language models have demonstrated remarkable progress in code generation, the generated code often exhibits poor runtime efficiency, limiting its practical application in performance-sensitive scenarios. To address this limitation, we propose an efficiency-oriented reinforcement learning framework guided by a novel performance reward. Based on this framework, we take a deeper dive into the code efficiency problem, identifying then proposing methods to overcome key bottlenecks: (1) Dynamic exploration overcomes the static data constraints of offline fine-tuning, enabling the discovery of more efficient code implementations. (2) The error-insensitive reinforcement learning method and high-contrast efficiency signals are crucial for mitigating systematic errors and achieving effective optimization. (3) Online exploration is most effective when starting from a high-correctness baseline, as this allows for efficiency improvements without sacrificing accuracy. With these discoveries, we finally propose a two-stage tuning method, which achieves high and balanced performance across correctness and efficiency. The results of experiments show the effectiveness of the method, which improves code correctness by 10.18\% and runtime efficiency by 7.75\% on a 7B model, achieving performance comparable to much larger model.

Although recent advances in quantum machine learning (QML) offer significant potential for enhancing generative models, particularly in molecular design, a large array of classical approaches still face challenges in achieving high fidelity and validity. In particular, the integration of QML with sequence-based tasks, such as Simplified Molecular Input Line Entry System (SMILES) string reconstruction, remains underexplored and usually suffers from fidelity degradation. In this work, we propose a hybrid quantum-classical architecture for SMILES reconstruction that integrates quantum encoding with classical sequence modeling to improve quantum fidelity and classical similarity. Our approach achieves a quantum fidelity of approximately 84% and a classical reconstruction similarity of 60%, surpassing existing quantum baselines. Our work lays a promising foundation for future QML applications, striking a balance between expressive quantum representations and classical sequence models and catalyzing broader research on quantum-aware sequence models for molecular and drug discovery.