TREE NEWS update: SenseTime officially launched the SenseNova Loop-9B model, which carries roughly 9 billion total parameters. The model expands 32 physical decoding layers into 56 layer computations by executing shared intermediate layers twice, and supports text and image input with text output at a context length of up to 512K tokens.
SenseTime Releases SenseNova Loop-9B Model with 512K Context
The interesting detail is architectural rather than the headline launch: reusing shared intermediate layers to expand effective depth suggests SenseTime is optimizing for capability per parameter, which matters for deployment economics. The 512K context and multimodal input place it in the long-context segment where enterprise document and image-heavy workflows compete. Whether the layer-looping approach translates into real-world gains beyond benchmarks is the open question, and how it is received by developers evaluating open-weight alternatives is worth watching.
Generated by AI for reference only.
Share on WeChat
Open WeChat → Scan → then tap "…" to send to a chat or Moments.
Tap "…" in the top-right corner to send to a chat or share to Moments.