Create

Sign in to ReadmeX

or
ReadmeX
ReadmeX

A clearer picture, in a conversation.

Catch up on what matters, then ask a little deeper.

Your community and people briefings stay personal to you.

ByteDance Seed

AI overview

Sign in and the AI will write an overview from our coverage.

Headlines · 3

  1. ByteDance Seed says it found cause of DeepSeek's long-context performance drift

    According to a DoNews report, ByteDance's Seed team says it has identified the cause of DeepSeek models "acting up" (producing erratic outputs), pointing to possible performance drift in long-context scenarios. The report carries headline-level information only, with no model versions, test methods or detailed conclusions given, and no public response from ByteDance or DeepSeek.

    DoNews · 🔥 9
  2. ByteDance Seed paper ties DeepSeek long-context swings to KV-cache phase sensitivity

    Per DoNews, ByteDance's Seed team posted an arXiv paper in late September 2026 arguing that "phase sensitivity" introduced by chunked KV-cache compression is the main driver of long-context retrieval fluctuations in DeepSeek-V4 series models. Covering open-weight models including DeepSeek-V4-Flash, V4-Pro and V4.1-Flash, the work says compression windows impose token phase coordinates, so the same information can be retrieved with accuracy differing by as much as 40 percentage points across phases — a periodic degradation that averaged benchmark scores tend to hide. The team calls for better cache-compression designs to improve long-context stability.

    DoNews · 🔥 9
  3. Alibaba said to lead $300M round in AI data startup UniPat at $2.5B valuation

    Foreign media reported in September that Alibaba plans to lead a $300M funding round in AI training and evaluation company UniPat at roughly a $2.5B valuation, with Tencent and Sequoia China reportedly also interested. UniPat was founded by Li Kuan, previously at Alibaba's Tongyi AI Lab working on post-training, data synthesis and reinforcement learning, while CTO Chen Liang worked on Qwen and Moonshot's multimodal teams. The piece also describes a wave of similar startups founded by ex-model-team staff, including Copula Lab (ex-MiniMax) and Zhineng Zhishi (ex-ByteDance Seed), building post-training data, agent environments and benchmarks.

    品玩 实时要闻 · 🔥 4

Experience and discussion from the community

Share my ByteDance Seed experienceAsk about ByteDance Seed

Nobody has shared their experience with ByteDance Seed yet.