A Quest After Perspectives

🧬 Overview G-LNS (Generative Large Neighborhood Search) represents a breakthrough in automated algorithm design, leveraging Large Language Models to automatically create Large Neighborhood Search operators for combinatorial optimization problems. Unlike traditional approaches that restrict designs to fixed heuristic forms, G-LNS enables structural algorithmic innovation through the co-evolution o…

Highlights First framework to co-evolve destroy and repair operators for Large Neighborhood Search using LLMs Synergy Matrix explicitly models operator interactions during evolutionary process Dual-population architecture maintains separate populations for destroy and repair operators Generative design produces executable code rather than just parameter tuning Strong generalization to unseen prob…

Upcoming challenges such as MLGWSC2, currently at the proposal stage, provide a new testbed for exploring machine-learning–based approaches to gravitational-wave analysis. In this flash talk, I briefly introduce my core ideas and experience using evolutionary algorithms, Evo-MCTS, and reinforcement learning as adaptive search and optimization tools. I outline key methodological insights and discu…

AI x Cosmology: From Computational Tools to Scientific Discovery Exploring the transition from AI as a computational tool to new paradigms in scientific discovery within cosmology.

Fantasy, Reality, and the Cost of Becoming a Graduate Student Mindset, Skills, and the Unwritten Rules of Graduate Life | 去魅之后的研究生之路：觉悟、技巧与“人情世故“

Interpretable Gravitational Wave Data Analysis with Reinforcement Learning and Large Language Models MLA Call (2025/12/18) Webnier 23:00-23:15. Based on 2024 Mach. Learn.: Sci. Technol. 5 015046 (arxiv: 2212.14283)

Interpretable Gravitational Wave Data Analysis with Reinforcement Learning and Large Language Models Based on arXiv:2508.03661

Abstract Presented at FQCP2025 on 2025-11-13. See event page: https://indico.ictp-ap.ucas.ac.cn/event/4/ Location 杭高院 Hangzhou, Zhejiang

Interpretable Gravitational Wave Data Analysis with Reinforcement Learning and Large Language Models Based on arXiv:2508.03661

Interpretable Gravitational Wave Data Analysis with Reinforcement Learning and Large Language Models Based on arXiv:2508.03661

Abstract Based on arXiv:2508.03661 Date Sep 10, 2025 7:00 AM — 7:20 AM

Gravitational Wave Data Analysis with Reinforcement Learning and Large Language Models Based on arXiv:2508.03661

Highlights Breakthrough in Automated Algorithm Discovery : Evo-MCTS represents a paradigm shift in scientific computing by enabling automated discovery of interpretable algorithms that match or exceed human-designed solutions. Exceptional Performance Gains : Achieves 20.2% improvement over domain-specific methods and 59.1% improvement over LLM-based optimization frameworks in gravitational wave d…

🧬 Overview Evo-MCTS represents a breakthrough in automated scientific algorithm discovery, introducing the first integration of Large Language Model (LLM) guidance with domain-aware physical constraints for gravitational wave detection. This groundbreaking framework systematically explores algorithmic solution spaces through tree-structured search enhanced by evolutionary optimization, addressing…

Highlights Comprehensive Survey : First comprehensive review of simulation-based inference (SBI) methods specifically tailored for gravitational wave data analysis, covering both theoretical foundations and practical applications. Five Major SBI Frameworks : In-depth coverage of Neural Posterior Estimation (NPE), Neural Ratio Estimation (NRE), Neural Likelihood Estimation (NLE), Flow Matching Pos…

Interpretable Gravitational Wave Data Analysis with Deep Learning and Large Language Models

原：AI赋能信号处理：引力波与通信系统案例分析

Date Apr 20, 2025 10:30 AM — 11:00 AM

Date Apr 8, 2025 10:00 AM — 11:30 AM