Loading the edition…
· Past edition · 3 stories
Retrospective edition reconstructed from later captures. Paper versions may include later updates; coverage labels describe the evidence used.
Research & academia
A full-duplex system separates live dialogue from background tool work.
Why it matters: Live speech interaction needs to continue while background tools and reasoning run.
Research & academia
A training system targets several separate memory peaks in large MoE models.
Open each story in the reader. Summaries are shown when available; some sources provide an abstract only. With an account you can keep your own notes.
Why it matters: Long-context MoE training must address several distinct memory peaks, not only the largest one.
Research & academia
A desktop serving benchmark measures memory and fidelity alongside speed.
Why it matters: Serving comparisons need memory and output fidelity alongside speed.