RPI researcher spanning generative models of neural spiking and vision-language-action robot learning.
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-09) 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—
This axis is unscored: the current snapshot covers only two 2026 preprints, so publication-volume data is insufficient. It would measure total research output and its consistency across years.
Impact—
No impact score is available: citation counts and downstream-uptake signals are absent from the current data. This axis would measure how much later work builds on these two preprints.
Focus0/100
The focus score is 0/100, driven by an exact 50/50 field distribution between computational neuroscience and robotics (vision-language-action models / robot learning) across the two cataloged papers. The record is dual-track: a discrete generative model for HD-MEA spiking on one side, visual masking against VLA trajectory overfitting on the other. Focus measures topical concentration, so this score flags breadth rather than weakness — both lines are fresh 2026 arXiv work.
Collaboration53/100
Collaboration scores 53/100 with 13 distinct coauthors in the snapshot. Authorship roles differ by line: last author on the neuronal-spiking generative-model paper, middle author on MaskVLA — consistent with a senior supervisory position in the neuroscience work and a contributor role in robotics. The recurring list spans twelve names, including Md Sayed Tanveer, Mohammed A. Mostajo-Radji, Yuxuan Jiang, Shenhao Yan, Qi Liu, Qing Zhao and Shuguang Cui.
Activity—
Activity is unscored: the snapshot has no multi-year publication cadence, only two 2026 preprints. This axis would measure publishing rhythm and recent research momentum over time.
Reproducibility78/100
Reproducibility scores 78/100, an average over two scored papers. The discrete neuronal-spiking model carries high reproducibility readiness, while MaskVLA is rated medium: Evolve notes strong RoboTwin 2.0 and real-robot gains but no error bars and an augmentation confound. The spiking paper's verdict likewise tempers its 5.2× tokenizer and 1.4–2.6× site-AP gains against a stronger per-assay lookup baseline.
Research summary
Ge Wang is a researcher at Rensselaer Polytechnic Institute whose cataloged work splits evenly (0.5 and 0.5) between computational neuroscience and robotics, specifically vision-language-action models and robot learning. The Evolve catalog holds two 2026 arXiv preprints. In the neuroscience line, Wang is the last author on a discrete generative model of neuronal spiking on microelectrode arrays, pairing a vector-quantized autoencoder with a masked generative transformer to encode ultra-sparse HD-MEA recordings, including human brain organoids, into a discrete motif alphabet; Evolve reports a 5.2× tokenizer gain and 1.4–2.6× site-AP gains, while noting that a per-assay lookup baseline still outranks it. In robotics, Wang is a middle author on MaskVLA, which applies primary-camera visual masking during VLA fine-tuning to counter trajectory overfitting, with gains on RoboTwin 2.0 and real robots, though without error bars. Radar scores collaboration at 53 with 13 distinct coauthors, focus at 0 from the even two-field split, and reproducibility at 78 averaged over two papers; output, impact and activity remain unscored for lack of data. Recurring coauthors include Md Sayed Tanveer, Mohammed A. Mostajo-Radji, Yuxuan Jiang, Shenhao Yan, Qi Liu, Qing Zhao and Shuguang Cui.
Collaboration pattern
The snapshot lists 13 distinct coauthors and a mid-range collaboration score of 53/100. Twelve recurring collaborators span both research lines: Md Sayed Tanveer, Mohammed A. Mostajo-Radji, Yuxuan Jiang, Jiaying Huang, Shenhao Yan, Jiahao Yang, Chengsi Yao, Qi Liu, Qing Zhao, Shuguang Cui, Yiming Zhao and Yatong Han. Authorship roles differ by field — last author on the neuronal-spiking model, middle author on MaskVLA — consistent with a senior position in the neuroscience line and a co-contributor role in robotics.
Research keywords
Highlights
Discrete motif alphabet for ultra-sparse HD-MEA spiking
A per-assay lookup baseline still leads the spiking tokenizer
MaskVLA: masking the primary camera to curb trajectory overfitting
Timeline
- 2026Last-author arXiv preprint: a discrete generative model of neuronal spiking activity on microelectrode arrays (computational neuroscience).
- 2026Middle-author arXiv preprint: MaskVLA, visual masking against trajectory overfitting of vision-language-action models (robotics).
- 2026Both preprints enter the Evolve catalog, averaging 78/100 on reproducibility and splitting evenly between the two fields (0.5/0.5).
Papers in the Evolve catalog
Frequent co-authors
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.
- Radar scores are computed from a small bibliometric snapshot covering only the two cataloged 2026 papers; treat all figures as provisional.
- Output, impact and activity axes are unscored because publication counts, citations and yearly publishing cadence data are not yet available.
- 'Ge Wang' is a common name; this profile is name-resolved from public bibliographic data and the two reviewed papers, without independent identity verification.
- No ORCID or country was provided; affiliation is stated as Rensselaer Polytechnic Institute and is not further verified.