Year: 2026
Type: preprint
Abstract: We present a suite of 50 unpublished research-level theoretical physics problems with verifiable answers. Without AI assistance, a typical problem in our collection could be solved by a world expert in about a day or less but would take considerably longer for experts in adjacent specialties. We test the ability of off-the-shelf AI agents to tackle these problems autonomously, without human hints. When given access to tools available in a realistic setting, including the ability to search the literature and write scripts for numerical and symbolic work, the best-performing agent, GPT-6-Astra harnessed in Codex, obtains the final answers with a near-perfect accuracy of 98% at an average cost under USD 3 per problem. On average, agents show only small changes in accuracy when denied web search and access to background research papers, suggesting significant memorization of the literature. These results suggest that state-of-the-art agents can accurately perform concrete calculations required across theoretical physics research. We comment on the broader implications of these findings and discuss possible explanations for the current lack of agent-led breakthroughs in theoretical physics despite their outstanding problem-solving abilities.
Source: arXiv (Cornell University)
Authors: Chandra, A. Ramesh, Mahajan, Aaghaz, Sanyal, Ainesh, Wang, Mianqi, Chakraborty, Shubhadeep, Chakraborty, Sudip, Chandrasekharan, Shailesh, Govindarajan, Rama, Goyal, Akshit, Heckman, Jonathan J., Iqbal, Yasir, Jaganathan, Divya, Kapoor, Saumav, Karch, Andreas, Khandare, Pushkar, Krishna, Sandeep, Kundu, Anupam, Laddha, Alok, Lin, Tongyan, Maity, Swarnendu, Mathur, Nilmani, Meynet, Shani Nadir, Mininno, Alessandro, Mishra, Abhijeet, Niranjan, Vidur, Prakash, Abhishodh, Rath, Pratik, Roy, Durbar, Sarkar, Debargha, Sastry, Srikanth, Shiu, Gary, Sinha, Aninda, Van Riet, Thomas, Netrapalli, Praneeth, Raju, Suvrat
Institutions: International Centre for Theoretical Sciences, Tata Institute of Fundamental Research, The University of Texas at Austin, Kavli Institute for the Physics and Mathematics of the Universe, The University of Tokyo, Homi Bhabha National Institute, Harish-Chandra Research Institute, Duke University, University of Pennsylvania, Indian Institute of Technology Madras, Harvard University, Jawaharlal Nehru Centre for Advanced Scientific Research, Chennai Mathematical Institute, University of California San Diego, University of Wisconsin–Madison, Monash University, University of Calgary, Indian Institute of Science Bangalore, KU Leuven
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7221814168