ByteDance paper finds periodic blind spots in DeepSeek V4 chunked KV-cache compression
A ByteDance team paper, “Periodic Weak Spots: Phase Sensitivity from Chunked KV-Cache Compression,” says the chunked KV-cache compression used by the DeepSeek V4 series creates “phase sensitivity”: adding only decorative equals signs ahead of a code snippet flipped DeepSeek-V4-Flash-Base's completion between 8 and 32, with an error rate of about 71.3% when the count of equals signs modulo 4 was 0 or 1 and about 91.5% accuracy when it was 2 or 3. In a 128K-token needle-in-a-haystack test with roughly 16,000 key-value pairs, moving a single pair changed accuracy by up to 40.2% for V4-Flash-Base and 34.8% for V4-Pro-Base, with gaps of 19.1% and 14.8% remaining after post-training; halving the compression stride in DeepSeek-V4.1-Flash narrowed the gap to 6.1% but did not remove it. Control models trained from scratch on the Qwen3-0.6B architecture showed the same periodic swings whenever fixed-length chunked compression was used, while a full-attention baseline did not, which the authors present as a structural flaw rather than a tuning issue.
For teams using long context for contract review or code completion, it is a reminder that chunked KV compression can cut cost while introducing position-dependent information loss.