pith. sign in
Pith Number

pith:MLZKOIP6

pith:2026:MLZKOIP6MHGCTDJRSGGMHSTZLE
not attested not anchored not stored refs pending

FiRe: Fine-grained Multimodal Reasoning for Enhanced Image Generation

Hyomin Kim, Jeeyoung Yun, Minjun Kim, Sungwoong Kim, Yerin Kim, Yongjin Kim, Yoonjin Oh, Yujung Heo

Decomposing text prompts into semantic units and verifying each via visual questions lets multimodal models refine generated images with targeted fixes.

arxiv:2604.13491 v3 · 2026-04-15 · cs.CV

Add to your LaTeX paper
\usepackage{pith}
\pithnumber{MLZKOIP6MHGCTDJRSGGMHSTZLE}

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

FiMR consistently outperforms image generation baselines, including reasoning-based methods, particularly on compositional text-to-image benchmarks.

C2weakest assumption

That VQA-based verification of decomposed prompt units produces reliable, unbiased fine-grained feedback that leads to targeted improvements without introducing new errors or hallucinations.

C3one line summary

FiMR improves text-to-image alignment by breaking prompts into minimal units, verifying each with VQA, and making localized refinements using MLLM reasoning.

Receipt and verification
First computed 2026-05-27T01:05:54.608818Z
Builder pith-number-builder-2026-05-17-v1
Signature Pith Ed25519 (pith-v1-2026-05) · public key
Schema pith-number/v1.0

Canonical hash

62f2a721fe61cc298d31918cc3ca795932f0413baad48cfb02a6db9a25f5e6d4

Aliases

arxiv: 2604.13491 · arxiv_version: 2604.13491v3 · doi: 10.48550/arxiv.2604.13491 · pith_short_12: MLZKOIP6MHGC · pith_short_16: MLZKOIP6MHGCTDJR · pith_short_8: MLZKOIP6
Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/MLZKOIP6MHGCTDJRSGGMHSTZLE \
  | 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: 62f2a721fe61cc298d31918cc3ca795932f0413baad48cfb02a6db9a25f5e6d4
Canonical record JSON
{
  "metadata": {
    "abstract_canon_sha256": "2a765cc5533c9d5e8ef9c699e63b199144bb2c2c0041a4a1cb6020dcc407bc50",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2026-04-15T05:24:29Z",
    "title_canon_sha256": "61fb2bfc827015116a5c93382df4b591306b087e102f7c60aabe079c7869bf13"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2604.13491",
    "kind": "arxiv",
    "version": 3
  }
}