February 2026

arXiv:2602.15832v1 Announce Type: new Abstract: Existing user simulations, where models generate user-like responses in dialogue, often lack verification that sufficient user personas are provided, questioning the validity of the simulations. To address this core concern, this work explores the task of identifying relevant but unknown personas of the simulation target for a given simulation context. We introduce PICQ, a novel dataset of context-aware choice questions, annotated with unknown personas (e.g., ”Is the user price-sensitive?”) that may influence user […]

Ver mais

arXiv:2602.15831v1 Announce Type: new Abstract: AI agents are increasingly deployed as autonomous systems capable of planning, tool use, and multi-agent collaboration across complex tasks. However, existing agent-related protocols focus on agent-to-agent interactions, leaving humans as external observers rather than integrated participants within the agent systems. This limitation arises from the lack of a standardized mechanism for agents to discover, address, and interact with humans across heterogeneous messaging platforms. In this paper, we propose the A2H (Agent-to-Human) protocol, a […]

Ver mais

The Early Days of Infrastructure Operations Imagine you’re the person responsible for managing and maintaining the servers for a company’s website or application. This used to be a simple job, but with technological advancements, infrastructure management has become more complex. In the early days, system administrators often logged into servers manually via SSH (Secure Shell) to configure servers, install software, or solve problems. SSH: The Initial Solution SSH changed everything in the 1990s. Before SSH, people had to […]

Ver mais

After my stint with Openclaw, I decided to build mine. While deciding how to move forward, I discovered Nanobot. It caught my attention immediately: only ~4,000 lines of Python versus OpenClaw’s 430,000+ lines of TypeScript. Clean, readable architecture. Easy to audit and extend. Actually designed for research and modification. Despite its elegance, it was still lacking in the security department. I decided to use it as a foundation and extend it myself. Enter Minion The name was inspired […]

Ver mais

When people talk about marketplaces, they usually talk about demand – growth, users, and marketing channels.  But in practice, marketplaces live or die by supply – and by how early you turn supply into a system rather than a manual operation. I learned this the hard way while building and scaling a two-sided marketplace in the local services space. Over several years, my team and I grew the platform from an early proof-of-concept to tens of thousands of […]

Ver mais

Table of Links Abstract and 1. Introduction Related Work Preliminaries and Notations Differentiable Structural Information 4.1. A New Formulation 4.2. Properties 4.3. Differentiability & Deep Graph Clustering LSEnet 5.1. Embedding Leaf Nodes 5.2. Learning Parent Nodes 5.3. Hyperbolic Partitioning Tree Experiments 6.1. Graph Clustering 6.2. Discussion on Structural Entropy Conclusion, Broader Impact, and References Appendix A. Proofs B. Hyperbolic Space C. Technical Details D. Additional Results D. Additional Results The hyperbolic partitioning trees of Cora is visualized in […]

Ver mais

Table of Links Abstract and 1. Introduction Related Work Preliminaries and Notations Differentiable Structural Information 4.1. A New Formulation 4.2. Properties 4.3. Differentiability & Deep Graph Clustering LSEnet 5.1. Embedding Leaf Nodes 5.2. Learning Parent Nodes 5.3. Hyperbolic Partitioning Tree Experiments 6.1. Graph Clustering 6.2. Discussion on Structural Entropy Conclusion, Broader Impact, and References Appendix A. Proofs B. Hyperbolic Space C. Technical Details D. Additional Results C. Technical Details C.1. Notations The mathematical notations are described in Table […]

Ver mais

Scientific discovery pipelines typically involve complex, rigid, and time-consuming processes, from data preparation to analyzing and interpreting findings. Recent advances in AI have the potential to transform such pipelines in a way that domain experts can focus on interpreting and understanding findings, rather than debugging rigid pipelines or manually annotating data. As part of an active collaboration between data science/AI researchers and behavioral neuroscientists, we showcase an example AI-enhanced pipeline, specifically designed to transform and accelerate the way […]

Ver mais