←Search results

By relevance · Total 93 results Clear search

Friday

Thursday

Tuesday

  1. Huxiu

    Carnegie Report: China Overtakes the US for the First Time as the Top Destination for Leading AI Talent, but Talent Still Nets Outflow

    A global AI talent report released by Carnegie China on September 23 shows that in 2025, 41% of the world's top AI researchers work in China and 34% in the US, compared with 27% versus 46% three years ago. Tracking the résumés of 10,280 authors from NeurIPS 2025, the report notes a net outflow of 1,729 people from China and a net inflow of 2,145 into the US; 98% of top researchers working in China did their undergraduate studies in China, versus only about 30% in the US. Among institutions, Peking University rose to first place, with Tsinghua second and Shanghai Jiao Tong University fourth. The report also acknowledges that its sample skews toward academia and underestimates industry.

    A global AI talent report released by Carnegie China on September 23 shows that in 2025, 41% of the world's top AI researchers work in China and 34% in the US, compared with 27% versus 46% three years ago. Tracking the résumés of 10,280 authors from NeurIPS 2025, the report notes a net outflow of 1,729 people from China and a net inflow of 2,145 into the US; 98% of top researchers working in China did their undergraduate studies in China, versus only about 30% in the US. Among institutions, Peki

    Matched fieldsBody

Monday

  1. Fireworks AI

    Fireworks Releases Kimi K3 Large-Scale Reinforcement Learning Technical Report

    Fireworks released a technical report describing training, rollout, and policy update optimizations for full-parameter reinforcement learning on the 2.8T-parameter Kimi K3. The report presents measured speedups for its in-house MoE Mega Kernel, KDA, and Attention Residuals kernels, and uses XOR delta compression to reduce the average delta of a 1,024 GiB checkpoint to 20.3 GiB. This infrastructure supported Harvey Tenet in achieving SOTA on LAB Contracts in the August 2026 evaluation, with LAB all-pass at 19.7%, 1.8 times that of base K3.

    Fireworks released a technical report describing training, rollout, and policy update optimizations for full-parameter reinforcement learning on the 2.8T-parameter Kimi K3. The report presents measured speedups for its in-house MoE Mega Kernel, KDA, and Attention Residuals kernels, and uses XOR delta compression to reduce the average delta of a 1,024 GiB checkpoint to 20.3 GiB. This infrastructure supported Harvey Tenet in achieving SOTA on LAB Contracts in the August 2026 evaluation, with LAB a

    Matched fieldsBody

Sunday

Saturday

Friday

Sunday

Saturday

Wednesday

Friday