Shuaiqiang Wang

—

Unclaimed · AI-structured profileOpenAlex

LLM-agent researcher spanning multimodal document generation and agentic visual RAG; middle author on 2026 arXiv work

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

—

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-09-20) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.

Research profile

Output—

No score yet: the snapshot lacks publication-count and venue data, covering only two reviewed 2026 arXiv preprints. This axis would measure total output volume and authorship roles once a fuller bibliographic record is linked.

Impact—

No score yet: citation counts, h-index, and downstream uptake are absent from the input data, so influence cannot be estimated. This axis would measure citations and community adoption of the two reviewed preprints.

Focus0/100

Scores 0/100, reflecting an even 50/50 split of the reviewed output between two distinct fields — natural language processing (multimodal document generation) and multimodal retrieval-augmented generation / vision-language agents. With only two papers in the snapshot, this low score signals breadth across adjacent LLM-agent topics rather than demonstrated specialization; one further paper in either area would materially change the measured concentration.

Collaboration53/100

Scores 53/100. The reviewed work involves 13 distinct coauthors, with recurring partners including Yucheng Shen, Jianmin Wu, Lingyong Yan, Dawei Yin, Runze Li, Yukun Zhao, Can Xu, Jiulong Wu, and Min Cao. On both 2026 papers Wang sits in a middle author position rather than first or corresponding, consistent with a contributor inside larger teams rather than a solo or lead driver — a moderately connected, not hub-level, pattern.

Activity—

No score yet: yearly publishing rhythm cannot be computed because only 2026 paper records are provided, with no multi-year history. This axis would measure publishing cadence and sustained research engagement across recent years.

Reproducibility38/100

Scores 38/100, averaged over the two reviewed papers (reproPapers: 2). The poster-generation work (PosterVisor) carries medium reproducibility readiness, while the agentic visual RAG work (SCORE) is rated low — Evolve flags SOTA claims resting on an LLM-judge metric without variance estimates and partly non-comparable baselines. Reproducibility is thus the weakest measurable axis in this profile.

Research summary

Shuaiqiang Wang is a researcher working at the intersection of large language model agents and multimodal information systems. The Evolve catalog indexes two 2026 arXiv preprints, split evenly across two fields: natural language processing (multimodal document generation) and multimodal retrieval-augmented generation / vision-language agents — an exact 50/50 distribution that also explains this profile's focus score of 0 on a 0–100 scale. The first thread, PosterVisor, tackles scientific poster generation by wrapping LLM output in frozen rubric contracts with repair-aware rechecking; Evolve judges the gains real but modest, with ablation trade-offs. The second thread, SCORE, makes evidence selection and consolidation an explicit, trainable stage of agentic visual RAG using GRPO-style reinforcement learning, and reports state-of-the-art results on three benchmarks — though Evolve notes the evaluation rests on an LLM-judge metric without variance estimates and partly non-comparable baselines. On both papers Wang holds a middle author position, not corresponding author, within a network of 13 distinct coauthors (collaboration score 53/100). Reproducibility, averaged over the two reviewed papers, scores 38 — the poster work rated medium readiness, the RAG work low. No institution, country, or ORCID is attached to this profile; it is name-resolved from public bibliographic data, and the output, impact, and activity axes remain unscored for lack of underlying data.

Collaboration pattern

Wang's two reviewed 2026 papers involve 13 distinct coauthors, with recurring names including Yucheng Shen, Jianmin Wu, Lingyong Yan, Dawei Yin, Runze Li, Yukun Zhao, Can Xu, Jiulong Wu, and Min Cao. On both papers Wang is a middle author and not corresponding, suggesting contribution within larger LLM-agent teams; the collaboration score is 53/100.

Research keywords

scientific poster generationLLM agentscontract-based controlsemantic-geometric contractrecursive verificationlayout generationvisual retrieval-augmented generationagentic RAGevidence selectionevidence consolidationvision-language modelsGRPO reinforcement learning

Highlights

SCORE: evidence selection becomes a trainable stage of agentic visual RAG
The preprint formalizes evidence selection and consolidation as an explicit, trainable stage of agentic visual RAG, optimized with GRPO reinforcement learning, and reports SOTA on three benchmarks; Evolve cautions that gains rest on an LLM-judge metric without variance estimates and partly non-comparable baselines, with low reproducibility readiness.
PosterVisor: persistent contract-based control for LLM poster generation
The preprint adds frozen rubric contracts and repair-aware recursive rechecking to LLM-based scientific poster generation; Evolve judges the gains real but modest, notes ablation trade-offs, and rates reproducibility readiness medium.

Timeline

2026Agentic visual RAG preprint (SCORE) appears on arXiv: evidence selection and consolidation become an explicit, trainable stage, with SOTA reported on three benchmarks.
2026Scientific poster generation preprint (PosterVisor) appears on arXiv: frozen rubric contracts plus repair-aware rechecking yield real but modest gains with ablation trade-offs.
2026Both indexed papers land in the same year, splitting the profile evenly 50/50 across its two fields and involving 13 distinct coauthors.

Papers in the Evolve catalog

Frequent co-authors

Yucheng Shen×2 Jianmin Wu×2 Lingyong Yan×2 Dawei Yin×2 Runze Li×1 Yukun Zhao×1 Can Xu×1 Jiulong Wu×1 Min Cao×1 Lingyong Yan×1 Shuaiqiang Wang×1 Jianmin WU×1

About this profile

This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.

Bibliometric scores (collaboration 53, focus 0, reproducibility 38) are computed by Evolve from public bibliometric data (OpenAlex-derived); the output, impact, and activity axes lack underlying data.
No institution, country, or ORCID is available; Shuaiqiang Wang is a name-resolved identity from public bibliographic records and may be ambiguous among researchers sharing the name.
Only two 2026 arXiv preprints are reviewed; the coauthor list shows name variants and duplicates (e.g., 'Jianmin WU' vs 'Jianmin Wu', and the profile name itself appears in the list), so collaboration counts carry entity-resolution noise.
Citation counts, venue history, and h-index are not present in the input data and are therefore never quoted in this profile.