❌

Normal view

Received — 9 January 2026 ⏭ Favourites

Sci-Reasoning: A Dataset Decoding AI Innovation Patterns

✇cs.AI, q-bio.NC updates on arXiv.org
By: Jiachen Liu · Maestro Harmon · Zechen Zhang
9 January 2026 at 13:00
arXiv:2601.04577v1 Announce Type: new Abstract: While AI innovation accelerates rapidly, the intellectual process behind breakthroughs -- how researchers identify gaps, synthesize prior work, and generate insights -- remains poorly understood. The lack of structured data on scientific reasoning hinders systematic analysis and development of AI research agents. We introduce Sci-Reasoning, the first dataset capturing the intellectual synthesis behind high-quality AI research. Using community-validated quality signals and an LLM-accelerated, human-verified pipeline, we trace Oral and Spotlight papers across NeurIPS, ICML, and ICLR (2023-2025) to its key predecessors, articulating specific reasoning links in a structured format. Our analysis identifies 15 distinct thinking patterns, with three dominant strategies accounting for 52.7%: Gap-Driven Reframing (24.2%), Cross-Domain Synthesis (18.0%), and Representation Shift (10.5%). The most powerful innovation recipes combine multiple patterns: Gap-Driven Reframing + Representation Shift, Cross-Domain Synthesis + Representation Shift, and Gap-Driven Reframing + Cross-Domain Synthesis. This dataset enables quantitative studies of scientific progress and provides structured reasoning trajectories for training the next generation AI research agents.
Received — 3 November 2025 ⏭ Favourites

Spatial-SSRL: Enhancing Spatial Understanding via Self-Supervised Reinforcement Learning

✇cs.AI, q-bio.NC updates on arXiv.org
By: Yuhong Liu · Beichen Zhang · Yuhang Zang · Yuhang Cao · Long Xing · Xiaoyi Dong · Haodong Duan · Dahua Lin · Jiaqi Wang
3 November 2025 at 13:00
arXiv:2510.27606v1 Announce Type: cross Abstract: Spatial understanding remains a weakness of Large Vision-Language Models (LVLMs). Existing supervised fine-tuning (SFT) and recent reinforcement learning with verifiable rewards (RLVR) pipelines depend on costly supervision, specialized tools, or constrained environments that limit scale. We introduce Spatial-SSRL, a self-supervised RL paradigm that derives verifiable signals directly from ordinary RGB or RGB-D images. Spatial-SSRL automatically formulates five pretext tasks that capture 2D and 3D spatial structure: shuffled patch reordering, flipped patch recognition, cropped patch inpainting, regional depth ordering, and relative 3D position prediction. These tasks provide ground-truth answers that are easy to verify and require no human or LVLM annotation. Training on our tasks substantially improves spatial reasoning while preserving general visual capabilities. On seven spatial understanding benchmarks in both image and video settings, Spatial-SSRL delivers average accuracy gains of 4.63% (3B) and 3.89% (7B) over the Qwen2.5-VL baselines. Our results show that simple, intrinsic supervision enables RLVR at scale and provides a practical route to stronger spatial intelligence in LVLMs.
Received — 29 October 2025 ⏭ Favourites

Multimodal 3D Genome Pre-training

✇cs.AI, q-bio.NC updates on arXiv.org
By: Minghao Yang · Pengteng Li · Yan Liang · Qianyi Cai · Zhihang Zheng · Shichen Zhang · Pengfei Zhang · Zhi-An Huang · Hui Xiong
29 October 2025 at 12:00
arXiv:2504.09060v2 Announce Type: replace-cross Abstract: Deep learning techniques have driven significant progress in various analytical tasks within 3D genomics in computational biology. However, a holistic understanding of 3D genomics knowledge remains underexplored. Here, we propose MIX-HIC, the first multimodal foundation model of 3D genome that integrates both 3D genome structure and epigenomic tracks, which obtains unified and comprehensive semantics. For accurate heterogeneous semantic fusion, we design the cross-modal interaction and mapping blocks for robust unified representation, yielding the accurate aggregation of 3D genome knowledge. Besides, we introduce the first large-scale dataset comprising over 1 million pairwise samples of Hi-C contact maps and epigenomic tracks for high-quality pre-training, enabling the exploration of functional implications in 3D genomics. Extensive experiments show that MIX-HIC can significantly surpass existing state-of-the-art methods in diverse downstream tasks. This work provides a valuable resource for advancing 3D genomics research.
Received — 25 May 2025 ⏭ Favourites
Received — 13 April 2025 ⏭ Favourites

Early Screening and Subtype Identification of High-Risk Lung Nodules via Breathprint by Graphene eNose Platform: A Large Cohort Study

✇Pulmonary nodule
By: Xingyu Zhu · Qiaofen Chen · Jiajing Sun · Lichen Zhang · Zhengwei Huang · Jingwei Xu · Haichuan Hu · Yuqi He · Zhao Chen · Xiaogang Ye · Xueyin Chen · Aotian Guo · Sheng Lu · Tao Shen · Jianmin Wu · Zhengfu He
7 April 2025 at 18:00

ACS Sens. 2025 Apr 25;10(4):3101-3111. doi: 10.1021/acssensors.5c00314. Epub 2025 Apr 7.

ABSTRACT

Early screening of individuals with high-risk lung nodules can significantly improve the prognosis of lung cancer patients, and accurate identification of lung nodule subtypes can provide guidance for medical treatment. Exhaled breath (EB) analysis via eNoses offers a quick and noninvasive approach, but current eNose technology lacks quality control and solid validation in large population studies. Herein, an eNose platform integrated with a metal ion-decorated graphene sensor array and a breath sampling accessory was established. EB samples from 427 healthy subjects and 2586 subjects with lung nodules, including various benign and malignant subtypes, were collected through the breath sampling accessory for quality control. The large-cohort clinical EB samples were analyzed by the eNose platform to acquire the cross-reactive resistance response. Breathprint analysis for high-risk lung nodules using SVM and age-matched training sets yielded strong and robust performance. Combined with baseline data, the model achieved an AUC of 0.93 (95% CI, 0.89-0.96) on the external test set, with 97% sensitivity and 73% specificity. Moreover, dimensionality reduction analysis of breathprints demonstrated separability across different lung nodule subtypes. This study demonstrates the reliability of the graphene eNose platform to identify high-risk lung nodules and classify lung nodule subtypes in a noninvasive and rapid method.

PMID:40193324 | DOI:10.1021/acssensors.5c00314

There are no more articles
❌