❌

Normal view

Received — 10 September 2026 ⏭

Discovery of synthetic TBK1 activator inducing type I interferon-mediated antiviral and antitumor immunity

✇Molecular Therapy
By: Kwiwan Jeong · Subin Hong · Pin Liu · Jinbeom Seo · Taegwan Yun · Sun-mi Park · Myung Jin Song · Jung Eun Yang · Eun-Hye Jeong · Byungil Kim · Sungmin Song · Jieun Lee · Kyuhee Park · Su Yeon Lim · Jung Sang Park · Pan Soo Kim · Do-Geun Kim · Beom Soo Shin · Ji Hoon Jeong · Wonsik Lee
25 August 2026 at 08:00
Lee and colleagues identify MFT251 as a synthetic small-molecule TBK1 agonist that directly activates type I interferon signaling. MFT251 induces broad antiviral defenses and reshapes the tumor microenvironment to enhance antitumor immunity, establishing pharmacologic TBK1 activation as a strategy for innate immune modulation in infection and cancer.
Received — 27 May 2026 ⏭

Inverting the Shield: Systematically Generating Safety Tests from Policy Specifications

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xiaoyue Lu · Xianglin Yang · Haijun Liu · Jiahao Liu · Kuntai Cai · Yan Xiao · Jin Song Dong
26 May 2026 at 12:00
arXiv:2605.24883v1 Announce Type: new Abstract: The widespread integration of Large Language Models (LLMs) necessitates rigorous and systematic safety evaluation. Existing paradigms either rely on constructed benchmarks to assess safety from predefined perspectives, or employ dynamic red-teaming to probe potential vulnerabilities. While effective, these approaches face challenges, as they depend heavily on expert domain knowledge, offer limited systematic guarantees, and are vulnerable to rapid obsolescence. To address these limitations, we introduce a novel framework POLARIS that brings the rigor of specification-based software testing to AI safety. POLARIS first compiles unstructured natural-language policies into First-Order Logic (FOL) representations, establishing a traceable link between high-level rules and concrete test cases. This formalization enables the construction of a Semantic Policy Graph, where complex policy violation scenarios are encoded as traversable paths. By systematically exploring this graph, POLARIS uncovers compositional violation patterns, which are then instantiated into executable natural-language test queries, enabling coverage-driven and reproducible safety testing. Experiments demonstrate that POLARIS achieves higher policy coverage and attack success counts compared to established baselines. Crucially, by bridging formal methods and AI safety, POLARIS provides a principled, automated approach to ensuring LLMs adhere to safety-critical policies with verifiable traceability. We release our code at https://github.com/huac-lxy/POLARIS.

Uncertainty Reasoning with Large Language Models for Explainable Disease Diagnosis

✇cs.AI, q-bio.NC updates on arXiv.org
By: Xiaoyang Fan · Yufan Cai · Zhe Hou · Jin Song Dong
26 May 2026 at 12:00
arXiv:2605.25566v1 Announce Type: new Abstract: Clinical decision-making requires reasoning over incomplete, imprecise, and linguistically expressed patient narratives. While large language models (LLMs) excel at extracting latent information from natural language, they lack the verifiability and interpretability essential for trustworthy medical AI. We propose a neuro-symbolic reasoning framework that aligns LLMs with formal logic to enable explainable and formally verifiable medical diagnosis. Patient descriptions and clinical guidelines are embedded into a neural knowledge base, where LLMs extract structured medical entities, temporal relations, and fuzzy symptom patterns, which are decoded into a symbolic knowledge base expressed in fuzzy logic and declarative rules. We perform two-stage reasoning: (1) inductive symbolic generalization to capture diagnostic patterns from encoded narratives, and (2) inference verification via a logic programming engine to derive and validate diagnoses consistent with clinical standards. Each symptom is treated as a fuzzy predicate with probabilistic weights, and inference paths are auditable, adjustable, and compatible with physician feedback. Unlike purely statistical methods, our system supports iterative refinement: misalignment between LLM-generated diagnoses and ground truth can be traced, explained, and corrected through formal rules. By combining logic-based transparency, LLM adaptability, and probabilistic robustness, the framework enables human-aligned healthcare inference with strong generalization and verifiable, step-by-step reasoning chains. We validate our framework on public benchmarks, demonstrating effective reconciliation of symbolic reasoning and LLMs with real-world clinical narratives. Results show performance comparable to state-of-the-art LLMs, while additionally providing interpretable reasoning paths and formally verifiable diagnostic conclusions.

Recent advances in biosensing platforms utilizing exosomal biomarker profiling for cancer diagnosis

✇MRD
By: Sojin Song · Sujin Hyung · Jeeyun Lee · Hong Nam Kim · Nakwon Choi
20 May 2026 at 18:00

Biosens Bioelectron. 2026 May 15;309:118809. doi: 10.1016/j.bios.2026.118809. Online ahead of print.

ABSTRACT

Tumor-derived exosomes carry multidimensional molecular cargo, including surface proteins, microRNAs, and lipids that encode tumor identity and disease dynamics. These features support their application as biomarkers for liquid biopsy-based cancer diagnostics. Circulating tumor DNA undergoes rapid nuclease-mediated degradation, whereas exosomes retain stable molecular information that reflects the proteomic, transcriptomic, and metabolic states of parent tumor cells. However, clinical translation of exosome-based sensing remains limited by variability in isolation, biological heterogeneity, and the analytical difficulty of detecting low-abundance biomarkers in clinical samples. In this review, we examine cancer-specific exosomal signatures across breast, lung, colorectal, and gastric cancers and evaluate biosensing platforms for exosomal biomarker profiling. We integrate engineering principles, clinical performance metrics, and AI-assisted analysis across complementary biosensing modalities to establish a cross-platform analytical framework. We compare optical platforms based on surface plasmon resonance, localized surface plasmon resonance, and surface-enhanced Raman scattering with photoluminescence- and electrochemical-based platforms in terms of sensitivity, clinical compatibility, and translational potential. Furthermore, we examine artificial intelligence (AI)-assisted biosensing frameworks, including classical machine learning classifiers, deep convolutional networks, ensemble models, explainable AI methods, and large language model interfaces. We evaluate how each framework addresses high-dimensional spectral complexity, nonlinear relationships among signals, and inter-patient variability in exosomal data. Finally, we identify remaining challenges, such as the lack of standardized isolation protocols and the absence of large-scale clinical validation. We further highlight minimal residual disease monitoring and early-stage cancer detection as important and underexplored directions for AI-integrated exosomal biosensing in precision oncology.

PMID:42161118 | DOI:10.1016/j.bios.2026.118809

Received — 26 March 2026 ⏭

mSFT: Addressing Dataset Mixtures Overfitting Heterogeneously in Multi-task SFT

✇cs.AI, q-bio.NC updates on arXiv.org
By: Woosung Koh · Jeyoung Jeon · Youngjin Song · Yujin Cheon · Soowon Oh · Jaehyeong Choi · Se-Young Yun
25 March 2026 at 12:00
arXiv:2603.21606v3 Announce Type: replace-cross Abstract: Current language model training commonly applies multi-task Supervised Fine-Tuning (SFT) using a homogeneous compute budget across all sub-datasets. This approach is fundamentally sub-optimal: heterogeneous learning dynamics cause faster-learning tasks to overfit early while slower ones remain under-fitted. To address this, we introduce mSFT, an iterative, overfitting-aware search algorithm for multi-task data mixtures. mSFT trains the model on an active mixture, identifies and excludes the earliest overfitting sub-dataset, and reverts to that specific optimal checkpoint before continuing. Extensive evaluations demonstrate that mSFT consistently outperforms 4 baselines across 10 benchmarks and 6 base models. Further analysis confirms mSFT maintains robust gains across diverse dataset sizes, task granularities, and is insensitive to its single new hyperparameter (compute budget). Notably, at low compute budget, mSFT can improve performance while lowering training FLOPs. Ultimately, mSFT establishes a practical overfitting-aware algorithm for multi-task SFT that maximizes the potential of models across diverse data mixtures.
Received — 25 February 2026 ⏭

LLM-enabled Applications Require System-Level Threat Monitoring

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yedi Zhang · Haoyu Wang · Xianglin Yang · Jin Song Dong · Jun Sun
24 February 2026 at 13:00
arXiv:2602.19844v1 Announce Type: cross Abstract: LLM-enabled applications are rapidly reshaping the software ecosystem by using large language models as core reasoning components for complex task execution. This paradigm shift, however, introduces fundamentally new reliability challenges and significantly expands the security attack surface, due to the non-deterministic, learning-driven, and difficult-to-verify nature of LLM behavior. In light of these emerging and unavoidable safety challenges, we argue that such risks should be treated as expected operational conditions rather than exceptional events, necessitating a dedicated incident-response perspective. Consequently, the primary barrier to trustworthy deployment is not further improving model capability but establishing system-level threat monitoring mechanisms that can detect and contextualize security-relevant anomalies after deployment -- an aspect largely underexplored beyond testing or guardrail-based defenses. Accordingly, this position paper advocates systematic and comprehensive monitoring of security threats in LLM-enabled applications as a prerequisite for reliable operation and a foundation for dedicated incident-response frameworks.

Mantis: A Versatile Vision-Language-Action Model with Disentangled Visual Foresight

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yi Yang · Xueqi Li · Yiyang Chen · Jin Song · Yihan Wang · Zipeng Xiao · Jiadi Su · You Qiaoben · Pengfei Liu · Zhijie Deng
24 February 2026 at 13:00
arXiv:2511.16175v2 Announce Type: replace-cross Abstract: Recent advances in Vision-Language-Action (VLA) models demonstrate that visual signals can effectively complement sparse action supervisions. However, letting VLA directly predict high-dimensional visual states can distribute model capacity and incur prohibitive training cost, while compressing visual states into more compact supervisory signals inevitably incurs information bottlenecks. Moreover, existing methods often suffer from poor comprehension and reasoning capabilities due to the neglect of language supervision. This paper introduces Mantis, a novel framework featuring a Disentangled Visual Foresight (DVF) to tackle these issues. Specifically, Mantis decouples visual foresight prediction from the backbone with the combination of meta queries and a diffusion Transformer (DiT) head. With the current visual state provided to the DiT via a residual connection, a simple next-state prediction objective enables the meta queries to automatically capture the latent actions that delineate the visual trajectory, and hence boost the learning of explicit actions. The disentanglement reduces the burden of the VLA backbone, enabling it to maintain comprehension and reasoning capabilities through language supervision. Empirically, pretrained on human manipulation videos, robot demonstrations, and image-text pairs, Mantis achieves a 96.7% success rate on LIBERO benchmark after fine-tuning, surpassing powerful baselines while exhibiting high convergence speed. Real-world evaluations show that Mantis outperforms $\pi_{0.5}$, a leading open-source VLA model, particularly in instruction-following capability, generalization to unseen instructions, and reasoning ability. Code and weights are released to support the open-source community.
There are no more articles
❌