Tag
Shared an open collaborative repository Awesome Vibe Research maintained by ModelScope. This repository collects and curates reusable, verifiable, and evolvable AI-assisted components across the full research workflow, including agents, skills, workflows, tools, and best practices. It aims to help researchers and developers leverage AI to improve research efficiency.
This paper proposes that reliability in AI-assisted social science research depends on decision architecture—how cognitive labor is divided between humans and machines. Through a pre-specified factorial experiment, the authors show that an unconstrained multi-agent baseline fails in 72% of runs, while one organized with three architectural commitments (LLMs restricted to reasoning, deterministic data/estimation, and three human decision gates) fails in only 16%.
Researchers from Charles University introduce Bolzano, an open-source multi-agent LLM system that orchestrates prover and verifier agents to assist with mathematical research, reporting new results on six problems where four reached publishable quality and three were produced essentially autonomously.
GPT-5.2 assisted in deriving a new theoretical physics result showing that single-minus gluon tree amplitudes can be nonzero under specific half-collinear momentum conditions, challenging decades of assumptions in particle physics. The AI model identified patterns in complex Feynman diagram expressions and conjectured a general formula that was subsequently verified through formal proofs.