Continual-learning researcher bridging brain-inspired modular architectures and embodied multi-view video perception
2
papers here
2
reviewed by Evolve
2
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-13) 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—
Output is not scored: publication-volume data such as total paper counts, venues, and yearly productivity is not available in this snapshot. This axis would measure how much peer-reviewed and preprint work the researcher has produced over time.
Impact—
Impact is not scored: citation counts, h-index, and downstream adoption metrics are absent from the available data. This axis would measure how widely the cataloged papers are cited and used by the continual learning and computer vision communities.
Focus0/100
Focus scores 0/100, the lowest value on this axis, because the cataloged output splits evenly across two distinct subfields: machine learning (neuroAI / continual learning) and computer vision / machine learning (continual learning for embodied multi-view video perception), each carrying 0.5 of the field distribution. A 50/50 split between brain-inspired architectures and embodied video benchmarks reads as an exploratory, two-track agenda rather than a narrowed specialty at this stage.
Collaboration56/100
Collaboration scores 56/100 based on 16 distinct co-authors across the cataloged work. Both cataloged papers are first-authored (non-corresponding), so the pattern suggests steady team-based research rather than an isolated solo effort. Frequent co-authors include Kanglei Zhou, Yi Zhong, Liyuan Wang, Yuchen Liu, Qingyu Shi, Qi Cheng, Weiyi Dong, Chunyan Lan, Guanglong Sun, Jun Zhou, and Qian Li. The mid-range score reflects moderate breadth, not yet a large or leading network.
Activity—
Activity is not scored: recent publication cadence and year-by-year output trends cannot be computed from the current snapshot, which contains papers from 2026 only. This axis would measure the researcher's ongoing publishing rhythm and research momentum over recent years.
Reproducibility55/100
Reproducibility scores 55/100, computed as the mean readiness of the 2 reviewed papers (reproMean 55, reproPapers 2). Both papers are rated medium readiness: the brain-inspired modular continual learning paper delivers strong empirics but informal theory, and CE4L is judged largely convincing with router overclaim and seed variance flagged. A medium baseline suggests replication is plausible but not yet fully turnkey for independent teams.
Research summary
Hongwei Yan is an emerging researcher at the Center for Life Sciences working at the intersection of neuroscience-inspired machine learning and embodied computer vision, with two first-author arXiv preprints (2026) reviewed in the Evolve catalog. The first, "Brain-Inspired Hierarchical Modularity for General Continual Learning", draws on the Drosophila mushroom body to build a mixture-of-experts plus ensemble architecture; Evolve's verdict highlights large continual-learning gains in vision and robotics, strong empirics, informal theory, and competitiveness with replay baselines. The second, "CE4L: Continual Ego, Exo, and Ego-Exo Learning", contributes a continual multi-view video benchmark spanning egocentric, exocentric, and ego-exo settings, together with training-free subspace-routed parameter-efficient adapters; it is judged largely convincing, with router overclaim and seed variance flagged as open issues. The field distribution splits evenly (0.5/0.5) between neuroAI/continual learning and continual learning for embodied multi-view video perception, producing a focus score of 0/100 — a deliberate two-track agenda rather than a narrowed specialty. Collaboration scores 56/100 over 16 distinct co-authors including Kanglei Zhou, Yi Zhong, and Liyuan Wang, with both papers first-authored but not corresponding. Reproducibility averages 55/100 across 2 reviewed papers, each rated medium readiness. Output, impact, and activity are not yet scored because the underlying bibliometric data is unavailable.
Collaboration pattern
Across 16 distinct co-authors, Hongwei Yan's cataloged work is consistently first-authored but not corresponding, indicating a collaborative contributor within larger teams rather than a solo or leading-PI pattern. The recurring co-author group spans Kanglei Zhou, Yi Zhong, Liyuan Wang, Yuchen Liu, Qingyu Shi, Qi Cheng, Weiyi Dong, Chunyan Lan, Guanglong Sun, Jun Zhou, and Qian Li, suggesting a stable core team bridging machine learning and neuroscience-adjacent work.
Research keywords
Highlights
Brain-inspired hierarchical modularity delivers large continual-learning gains
CE4L benchmark extends continual learning to ego, exo, and ego-exo video
Timeline
- 2026Published first-author arXiv preprint "Brain-Inspired Hierarchical Modularity for General Continual Learning", the neuroAI/continual-learning track of the cataloged output.
- 2026Published first-author arXiv preprint "CE4L: Continual Ego, Exo, and Ego-Exo Learning", extending continual learning to embodied multi-view video perception.
- 2026Held first-author (non-corresponding) positions on both cataloged papers in the same year, with output split evenly between neuroAI/continual learning and embodied multi-view video perception (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.
- All scores are derived solely from the bibliometric snapshot behind this profile; the underlying database coverage is not specified in the input data.
- Output, impact, and activity axes are unscored because publication counts, citation metrics, and cadence data are missing from this snapshot.
- Reproducibility rests on only 2 reviewed papers, both 2026 arXiv preprints with medium readiness, so the 55/100 mean is a thin estimate.
- Country and ORCID are not provided and "Hongwei Yan" is a common name; this profile is name-resolved from public bibliographic data and the institution field only.