Skip to content
IFAI

字节跳动

Company28 stories

  1. Open sourceReach 48

    ByteDance open-sources BitDance to speed up AR image generation by 30x

    ByteDance just open-sourced BitDance, an autoregressive image generation model that accelerates high-resolution synthesis by over 30x. By replacing codebook indices with binary visual tokens and decoding patches in parallel via continuous-space diffusion, it achieves a record 1.24 FID on ImageNet. Why it matters: AR models typically suffer from slow generation speeds, but BitDance proves they can rival diffusion in efficiency.

  2. Open sourceReach 48

    ByteDance turns AI agent memory into a file system with OpenViking

    ByteDance just open-sourced OpenViking under Apache 2.0, introducing a virtual file system to manage agent contexts, memories, and skills instead of relying on flat vector slices. It uses tiered loading across three summary levels to slash token overhead while fully visualizing the agent's retrieval path.

  3. Generative mediaReach 41

    ByteDance previews Alive for synced audio-video generation

    ByteDance just unveiled Alive, a unified model that generates synchronized video and audio in one go from text prompts or reference images. By revamping diffusion architectures for tight cross-modal alignment, it claims to beat Sora in motion quality and lip-sync accuracy. While the research paper is out, the GitHub code and demo remain marked as coming soon.

  7. ResearchReach 36

    FlowAct-R1 streams lifelike AI avatars with infinite duration

    Researchers just introduced FlowAct-R1, a streaming humanoid video generation model capable of synthesizing infinite-duration video in real time. It achieves 25fps at 480p with a 1.5-second first-frame latency while maintaining natural behaviors. Why it matters: This drastically cuts latency for responsive AI avatars and seamless live streaming.

  8. ResearchReach 36

    ByteDance's SeedFold outpaces AlphaFold 3

    ByteDance's Seed team just introduced SeedFold, a biomolecular structure prediction model that beats AlphaFold 3 on key protein benchmarks. The team scaled the Pairformer width by 4x, introduced linear triangular attention to drop complexity to O(n²), and distilled 26.5M samples from AlphaFold 2. It marks a significant push forward in scaling laws for structural biology.