在学科竞赛领域，物理因题目复杂、推理强度高而长期被认为是人工智能（AI）最难攻克的挑战之一。与语言类任务相比，物理问题往往涉及图像识别、单位换算、公式推导和近似计算等多重环节，更考验系统是否具备对现实世界的理解与建模能力。

在近年来的科技发展中，人工智能（AI）逐渐成为各个领域的佼佼者。然而，物理学这一复杂且富有挑战性的学科，曾被认为是AI攻克的最难关卡之一。2025年国际物理奥林匹克竞赛中，名为Physics Supernova的AI系统以23.5分的佳绩，成功超越了人类金牌选手的平均得分，成为众人瞩目的焦点。

在学科竞赛领域，物理因题目复杂、推理强度高而长期被认为是人工智能（AI）最难攻克的挑战之一。与语言类任务相比，物理问题往往涉及图像识别、单位换算、公式推导和近似计算等多重环节，更考验系统是否具备对现实世界的理解与建模能力。随着AI日益深入现实世界，并不断迈向通用人工智能（AGI）乃至超级人工智能（ASI），能否通过物理抽象理解世界、解决问题，正在成为打造高水平智能系统的关键。

The supernova explosion, designated as SN 2018gv, is found in the spiral galaxy NGC 2525. According to NASA, Hubble began ...

What's the link between an exploding star, climate change and human evolution? Francis Thackeray, who has researched ancient ...

Researchers observed a supernova in progress that showed the onion-like structure of elements in stars, and the explosion may ...

In addition to revealing characteristics of mysterious particles called neutrinos, the massive DUNE experiment could help ...

In a recent groundbreaking discovery in astrophysics, astronomers have unveiled the presence of a “ghost star” nestled within ...

Around 11,300 years ago, a massive star teetered on the precipice of annihilation. It pulsed with energy as it expelled its ...