Pith. sign in
Pith Number

pith:ZJSJDOQX

pith:2026:ZJSJDOQXUFHGJCRLAUHOXEF3KP
not attested not anchored not stored refs pending

Making Image Editing Easier via Adaptive Task Reformulation with Agentic Executions

Bo Zhao, Haiyang Sun, Huan Yang, Kairui Guo, Kun Gai, Pengshan Wang, Runnan Du, Wei Ji, Yixin Cao

Reformulating vague image editing instructions into adaptive operation sequences with an MLLM agent lifts performance without changing the model.

arxiv:2604.15917 v2 · 2026-04-17 · cs.CV

Add to your LaTeX paper
\usepackage{pith}
\pithnumber{ZJSJDOQXUFHGJCRLAUHOXEF3KP}

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

Experiments on multiple benchmarks, including ImgEdit, PICA, and RePlan, across diverse editing backbones such as Qwen Image Edit and Nano Banana, show consistent improvements, with especially large gains on challenging cases.

C2weakest assumption

A large portion of these failures stem not from insufficient model capacity, but from poorly formulated editing tasks, such as those involving small targets, implicit spatial relations, or under-specified instructions.

C3one line summary

An MLLM agent reformulates image editing tasks into executable operation sequences to improve reliability on challenging cases across existing generative backbones.

Receipt and verification
First computed 2026-07-01T01:17:15.179822Z
Builder pith-number-builder-2026-05-17-v1
Signature Pith Ed25519 (pith-v1-2026-05) · public key
Schema pith-number/v1.0

Canonical hash

ca6491ba17a14e648a2b050eeb90bb53fe2487e85a7836094f06fb9450d2662b

Aliases

arxiv: 2604.15917 · arxiv_version: 2604.15917v2 · doi: 10.48550/arxiv.2604.15917 · pith_short_12: ZJSJDOQXUFHG · pith_short_16: ZJSJDOQXUFHGJCRL · pith_short_8: ZJSJDOQX
Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/ZJSJDOQXUFHGJCRLAUHOXEF3KP \
  | 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: ca6491ba17a14e648a2b050eeb90bb53fe2487e85a7836094f06fb9450d2662b
Canonical record JSON
{
  "metadata": {
    "abstract_canon_sha256": "c270b278e63cba776ddf084075a589312fde9f6e704df953090cfccd6be35a54",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2026-04-17T10:17:22Z",
    "title_canon_sha256": "ce1af3c70cd17f0b6af4df14a709b3e29b23e07dc5a25bb15302df67ce1dc1c0"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2604.15917",
    "kind": "arxiv",
    "version": 2
  }
}