Pith Number

pith:LTWHYP5B

pith:2026:LTWHYP5BK3AXMP4XSVVRIRNI6Q

not attested not anchored not stored refs pending

Towards On-Policy Data Evolution for Visual-Native Multimodal Deep Search Agents

Chenxin Li, Guanting Dong, Hangyu Guo, Hongru Wang, Junting Lu, Shijue Huang, Shuang Chen, Xinyu Geng, Yi R. Fung, Zhaochen Su, Zhenyu Li

On-policy data evolution from agent rollouts boosts multimodal deep search performance from 24.9% to 39% on average.

arxiv:2605.10832 v2 · 2026-05-11 · cs.CL

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

Add to your LaTeX paper

\usepackage{pith}
\pithnumber{LTWHYP5BK3AXMP4XSVVRIRNI6Q}

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

Across 8 multimodal deep search benchmarks, ODE improves the Qwen3-VL-8B agent from 24.9% to 39.0% on average, surpassing Gemini-2.5 Pro in standard agent-workflow setting (37.9%). At 30B, ODE raises the average score from 30.6% to 41.5%.

C2weakest assumption

That rollouts from the current policy produce training data that accurately identifies and fills the precise capability gaps without introducing self-reinforcing biases or training instability.

C3one line summary

A new image-bank harness and closed-loop on-policy data evolution method raises multimodal agent performance on visual search benchmarks from 24.9% to 39.0% for an 8B model and from 30.6% to 41.5% for a 30B model.

Receipt and verification

First computed	2026-06-08T01:04:07.308478Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

5cec7c3fa156c1763f97956b1445a8f42f43b136365004d76ce88b6d7247b0c5

Aliases

arxiv: 2605.10832 · arxiv_version: 2605.10832v2 · doi: 10.48550/arxiv.2605.10832 · pith_short_12: LTWHYP5BK3AX · pith_short_16: LTWHYP5BK3AXMP4X · pith_short_8: LTWHYP5B

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/LTWHYP5BK3AXMP4XSVVRIRNI6Q \
  | 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: 5cec7c3fa156c1763f97956b1445a8f42f43b136365004d76ce88b6d7247b0c5

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "f3a288909e3ddee0339012bde8bc17676df114c7ef398a2f4b48d6150606d5d6",
    "cross_cats_sorted": [],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2026-05-11T16:49:36Z",
    "title_canon_sha256": "e178d81e90584fcf37865ee23101edc46bb6f4f5cdd66e76965a755c80b267b4"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2605.10832",
    "kind": "arxiv",
    "version": 2
  }
}