AI researcher working on robot world-action models and LLM-agent exploration benchmarks, with two 2026 arXiv papers
2
papers here
0
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
0
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-28) 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 would score publication volume and consistency over time. The underlying record is too sparse to compute it: only a single 2026 entry with 2 works exists in the OpenAlex snapshot, so no multi-year output trend can be measured yet.
Impact—
Impact measures citation-based influence such as h-index and field-weighted citations. Here h-index and i10-index are both 0 and the two 2026 preprints are too recent to have accumulated citations, so this axis cannot be scored yet.
Focus—
Focus would measure topical concentration versus spread of research themes. Five OpenAlex topics each appear exactly once — from robot manipulation to AI ethics — a distribution too thin to distinguish deliberate focus from incidental topic assignment.
Collaboration68/100
The OpenAlex record shows 29 distinct co-authors around a core of 12 recurring collaborators (including Ming Zhang, Zhenghao Xiang, Peizhong Gao, Yujiong Shen and Yuhui Wang), supporting a collaboration score of 68/100. Both catalog papers place Ye in a middle author position on sizeable teams, indicating an active contributor inside multi-author groups rather than a solo or corresponding-lead pattern.
Activity32/100
Activity scores 32/100. The countsByYear profile contains a single entry — 2 works in 2026 — and recent3yWorks is likewise 2, so the visible output is concentrated entirely in the current year with no prior-year record in this snapshot. This yields a thin but fresh activity footprint: two arXiv preprints released in 2026, with nothing yet traceable before that window.
Reproducibility—
Reproducibility would score code, data, and protocol availability for independent replication. Evolve's paper-level reviews note medium readiness for Rolling-WAM and low for ExplorationBench, but no aggregate bibliometric signal exists yet from which to compute an axis score.
Research summary
Junjie Ye is an early-stage AI researcher whose visible publication record centers on two 2026 arXiv preprints reviewed in the Evolve catalog: Rolling-WAM, a world action model for robot manipulation that spreads video-action denoising across replanning steps via rolling diffusion and flow matching, and ExplorationBench, a benchmark that measures AI systems' exploration in verifiable alien worlds with exact grading. On both papers Ye is a middle-position author rather than first or corresponding, consistent with contributor roles inside larger teams. The OpenAlex record (identity resolved by name; no ORCID or institution provided) lists 2 works in the last three years — both in 2026 — 29 distinct co-authors, and five single-count topic assignments spanning robot manipulation and learning, GAN-based image synthesis, human pose and action recognition, ethics and social impacts of AI, and scientific computing and data management. Radar scores reflect this shape: collaboration 68/100 against a wide but young co-author network, and activity 32/100 with only 2 recent works. Citation indicators (h-index and i10-index of 0) simply show that 2026 preprints have not yet accumulated citations. Evolve's reviews flag strong engineering — 4.5x faster inference with competitive sim/real results on Rolling-WAM, and exact grading on ExplorationBench — alongside thin statistical validation: no significance tests and n=3 trajectories.
Collaboration pattern
Ye's collaboration footprint is wide and young: 29 distinct co-authors in the OpenAlex window with a stable core of 12 recurring collaborators, including Ming Zhang, Zhenghao Xiang, Peizhong Gao, Yujiong Shen, Yuhui Wang, Zhonghan Yue, Shihan Dou, Zhangyue Yin, Shichun Liu, Weihuang Zheng, Jiahao Chen, and Jiayi Chen. Both catalog papers are middle-author contributions to large teams, yielding a collaboration score of 68/100 without evidence of corresponding-author leadership.
Research keywords
Highlights
Rolling-WAM: 4.5x faster world-action inference via rolling diffusion
ExplorationBench: exact grading of AI exploration in alien worlds
Timeline
- 2026Rolling-WAM, a world action model with rolling imagination for robot manipulation, posted to arXiv (middle author).
- 2026ExplorationBench, a benchmark measuring AI systems' exploration in verifiable alien worlds, posted to arXiv (middle author).
- 2026OpenAlex snapshot records 2 works in 2026, 29 distinct co-authors, and 5 single-count topic assignments.
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.
- All bibliometrics come from a single OpenAlex record; the researcher is name-resolved from public bibliographic data, and 'Junjie Ye' is a common name, so identity confidence is limited.
- No ORCID, institution, or country was provided, so the profile contains no career biography and omits affiliation context entirely.
- Citation indicators are h-index 0 and i10-index 0, and both catalog papers are 2026 arXiv preprints — citation impact is simply not yet observable.
- Only 2 Evolve-reviewed papers underlie this profile, and the 5 topic assignments (1 count each) are too sparse for reliable field inference.