The Decoder· Maximilian Schreiner·· 17 h ago
GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece
GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece
AI summary
On the TheoPhysEval benchmark, GPT-6-Astra solves 49 of 50 unpublished theoretical physics calculations as accurately as leading specialists. In some cases where the agent seemed to fail, the experts' own reference solutions turned out to contain errors. But the researchers say physics hasn't had a breakthrough like the ones in mathematics, because the agents lack "taste."
Selection record
Threshold 76Media and individualsFirst 82Second 71
AdmittedSum of both 153 ≥ twice the threshold 152
- Source tier
- Media and individuals; this tier's threshold is 76
- Pre-filter
- passed：AI智能体解决物理研究问题，实质AI评测
- Why it was chosen
- Puts several AI agents through routine theoretical-physics calculations and reports the costs and failure cases, showing where agents actually stand on research computation.
A model scores each item twice, independently, against one written standard, out of 100. An item is admitted only when the two scores add up to twice the threshold. The threshold is set per source tier.
Source: The Decoder · the-decoder.com