[Submitted on 19 Mar 2025 (v1), last revised 17 Jul 2025 (this version, v2)] View a PDF of the paper titled Aligning Information Capacity Between Vision and Language via Dense-to-Sparse Feature Distillation for Image-Text Matching, by Yang Liu and 4 other authors View PDF HTML (experimental) Abstract:Enabling Visual Semantic Models to effectively handle multi-view description […]

[Submitted on 28 Oct 2024 (v1), last revised 16 Jul 2025 (this version, v3)] View a PDF of the paper titled SCULPT: Systematic Tuning of Long Prompts, by Shanu Kumar and 5 other authors View PDF Abstract:Prompt optimization is essential for effective utilization of large language models (LLMs) across diverse tasks. While existing optimization methods

[Submitted on 16 Jul 2025 (v1), last revised 17 Jul 2025 (this version, v2)] Authors:Yexuan Shi, Mingyu Wang, Yunxiang Cao, Hongjie Lai, Junjian Lan, Xin Han, Yu Wang, Jie Geng, Zhenan Li, Zihao Xia, Xiang Chen, Chen Li, Jian Xu, Wenbo Duan, Yuanshuo Zhu View a PDF of the paper titled Aime: Towards Fully-Autonomous Multi-Agent

arXiv:2507.13222v1 Announce Type: cross Abstract: A central question in computer science and statistics is whether efficient algorithms can achieve the information-theoretic limits of statistical problems. Many computational-statistical tradeoffs have been shown under average-case assumptions, but since statistical problems are average-case in nature, it has been a challenge to base them on standard worst-case assumptions. In

arXiv:2507.13090v1 Announce Type: cross Abstract: Robust XAI techniques should ideally be simultaneously deterministic, model agnostic, and guaranteed to converge. We propose MULTIDIMENSIONAL PROBLEM AGNOSTIC EXPLAINABLE AI (MUPAX), a deterministic, model agnostic explainability technique, with guaranteed convergency. MUPAX measure theoretic formulation gives principled feature importance attribution through structured perturbation analysis that discovers inherent input patterns and

I’m currently investigating ketamine, with the goal of assessing the risks of chronic use. For reasons I will get into in the real post, this is going to rely mostly on in vitro data, at least for neural damage, which means I need a way to translate real-world dosages into the concentration of ketamine in

From assuming oversight for complex workflows, such as procurement or recruitment, to carrying out proactive cybersecurity checks or automating support, enterprises are abuzz at the potential use cases for agentic AI.  According to one Capgemini survey, 50% of business executives are set to invest in and implement AI agents in their organizations in 2025, up

Transferring learned features to ultra-wideband radar We then ran a study that collected UWB radar data, along with electrocardiogram (ECG) and photoplethysmogram (PPG) data as our ground truth for heart rate, using a setup that placed the UWB radar sensor in positions where users typically hold their phone, i.e., on a table in front of

We should all be aware of the growing electricity toll of air-conditioning, but the AC hate is misplaced. Yes, AC is energy intensive, but so is heating our homes, something that’s rarely decried in the same way that cooling is. Both are tools for comfort and, more important, for safety.  So why is air-conditioning cast

The challenge of estimating an earthquake’s power One of the trickiest parts of an EEW system is estimating the magnitude of an earthquake in real-time. The magnitude tells us how big the earthquake is, which in turn determines how far the shaking will travel and who needs to be alerted. Getting this right is crucial