Pith Number

pith:6WDVKKSF

pith:2026:6WDVKKSFBK4XGFSUC5SCV4BXMY

not attested not anchored not stored refs resolved

Pyramid Forcing: Head-Aware Pyramid KV Cache Policy for High-Quality Long Video Generation

Guojie Luo, Jiawei Yang, Jiayi Luo, Jiayu Chen, Junbei Tang, Maoliang Li, Wenbiao Zhao, Xiang Chen, Zihao Zheng

Pyramid Forcing assigns different KV cache lengths to three attention head types to reduce error accumulation in long autoregressive video generation.

arxiv:2605.13111 v1 · 2026-05-13 · cs.CV

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{6WDVKKSFBK4XGFSUC5SCV4BXMY}

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

Pyramid Forcing consistently improves long-horizon generation quality on VBench-Long, increasing the 60-second Self Forcing score from 77.87 to 81.21 while enhancing motion dynamics, visual fidelity, and semantic consistency.

C2weakest assumption

The three head types (Anchor, Wave, Veil) are stable across different models and datasets and can be reliably identified offline without retraining or runtime overhead.

C3one line summary

Pyramid Forcing classifies attention heads into Anchor, Wave, and Veil types and applies type-specific KV cache policies to improve long-horizon autoregressive video generation quality.

References

37 extracted · 37 resolved · 10 Pith anchors

[1] Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion 2025 · arXiv:2506.08009

[2] MAGI-1: Autoregressive Video Generation at Scale 2025 · arXiv:2505.13211

[3] Causal forcing: Autoregressive diffusion distillation done right for high-quality real- time interactive video generation 2026

[4] Scalable diffusion models with transformers 2023

[5] Wan: Open and Advanced Large-Scale Video Generative Models 2025 · arXiv:2503.20314

Formal links

2 machine-checked theorem links

Receipt and verification

First computed	2026-05-18T03:08:58.115445Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

f587552a450ab973165417642af0376610e9f47f9fca2987620064cce90f9923

Aliases

arxiv: 2605.13111 · arxiv_version: 2605.13111v1 · doi: 10.48550/arxiv.2605.13111 · pith_short_12: 6WDVKKSFBK4X · pith_short_16: 6WDVKKSFBK4XGFSU · pith_short_8: 6WDVKKSF

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/6WDVKKSFBK4XGFSUC5SCV4BXMY \
  | 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: f587552a450ab973165417642af0376610e9f47f9fca2987620064cce90f9923

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "75348fc516cdb834c97cc03a6a77d15ec404b5a3b0f3efe1bd8c934986d90c1c",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2026-05-13T07:23:02Z",
    "title_canon_sha256": "ce189f10a6b255df28ffdde68e05f4dfca67e591ce6a0c610c34cf63fb17e514"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.13111",
    "kind": "arxiv",
    "version": 1
  }
}