ត្រឡប់ទៅព័ត៌មានវិញ

AI · អត្ថបទ

OpenAI Unveils MedChemBench to Test AI on Drug Chemistry Reasoning

Unite.AI ·

មើលត្រួសៗ

OpenAI released MedChemBench, a benchmark that tests whether AI models can handle drug discovery chemistry tasks like explaining molecules, comparing compounds, and predicting properties. GPT-6.2 Astra scored 84.5% of the available points at its highest reasoning setting. OpenAI says the scores alone do not prove that a model can actually help discover new drugs.

ចំណុចសំខាន់ៗ

ហេតុអ្វីវាសំខាន់

It shows how well AI handles drug-discovery chemistry, but OpenAI says it does not prove real drug-discovery value.

ប្រភពដើម

OpenAI, 의약화학 추론 평가용 MedChemBench 공개

Unite.AI · Nov 21, 04:20 KST

Read on Unite.AI ↗

AI summarized the original article. Check the original for accuracy. Copyright belongs to Unite.AI.

ការផ្សាយផ្សេងទៀត

OpenAI · ចំណាត់ថ្នាក់ និងព័ត៌មានបន្ថែម