Pith Number

pith:EQWLRQYM

pith:2025:EQWLRQYMRUPFS4KJGFZVGM247D

not attested not anchored not stored refs pending

Make It Long, Keep It Fast: End-to-End 10K Long User Behavior Sequence Modeling for Billion-Scale Douyin Recommendation

Beichuan Zhang, Bo Sun, Feng Zhang, Hangyu Wang, Jia-Qi Yang, Jinan Ni, Lin Guan, Qiwei Chen, Xiaowen Li, Xiao Yang, Xuanyuan Luo, Yi Cheng, Yuhang Qi, Zhifang Fan, Zhishan Zhao

Stacked cross-attention and shared batching let recommendation models process 10,000-length user histories at production speed.

arxiv:2511.06077 v3 · 2025-11-08 · cs.LG · cs.IR

Open paper page JSON Open Graph Bundle Merged state Verified badge What is a Pith Number?

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{EQWLRQYMRUPFS4KJGFZVGM247D}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

✓ Portable graph bundle live · download bundle · merged state

The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

Deployed at full traffic on Douyin, our system delivers significant improvements on key engagement metrics while meeting production latency, demonstrating a practical path to scaling end-to-end long-sequence recommendation to the 10k regime.

C2weakest assumption

That the length-extrapolative training strategy (train on shorter windows, infer on 10k) generalizes without performance loss and that stacked target-to-history cross attention captures the necessary preference signals from long histories.

C3one line summary

Douyin deploys stacked target-to-history cross attention and request-level batching to scale end-to-end recommendation modeling to 10k-length histories, observing scaling-law gains and live engagement improvements.

Formal links

2 machine-checked theorem links

Cited by

4 papers in Pith

One Pool, Two Caches: Adaptive HBM Partitioning for Accelerating Generative Recommender Serving

IAT: Instance-As-Token Compression for Historical User Sequence Modeling in Industrial Recommender Systems

One Pool, Two Caches: Adaptive HBM Partitioning for Accelerating Generative Recommender Serving

Receipt and verification

First computed	2026-05-20T01:05:01.391006Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

242cb8c30c8d1e597149317353335cf8c7c5156c744806cde9937756293981e5

Aliases

arxiv: 2511.06077 · arxiv_version: 2511.06077v3 · doi: 10.48550/arxiv.2511.06077 · pith_short_12: EQWLRQYMRUPF · pith_short_16: EQWLRQYMRUPFS4KJ · pith_short_8: EQWLRQYM

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/EQWLRQYMRUPFS4KJGFZVGM247D \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 242cb8c30c8d1e597149317353335cf8c7c5156c744806cde9937756293981e5

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "b4d0db21bf900b36037cdd67e71cb0bbb313470c8197f8e8cced64b2cfe73502",
    "cross_cats_sorted": [
      "cs.IR"
    ],
    "license": "http://creativecommons.org/licenses/by-sa/4.0/",
    "primary_cat": "cs.LG",
    "submitted_at": "2025-11-08T17:22:54Z",
    "title_canon_sha256": "cc635af1a9f45ca110d27edc47511eaba70be6c0f98b8bd6f75e285217ea6697"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2511.06077",
    "kind": "arxiv",
    "version": 3
  }
}