Robot learning and 3D vision researcher; last author on transferable 3D dynamics and efficient VLA adaptation work
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-09-19) 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—
Not yet scored: the snapshot contains no publication counts, venue history, or career-length output record for this researcher. This axis would measure total scholarly production over time — papers per year, venue tiers, and authorship roles across the full publication record.
Impact—
Not yet scored: no citation counts, h-index, or influence metrics are available in this snapshot. This axis would measure how often the researcher's work is cited and its uptake across robotics, computer vision, and machine learning communities.
Focus0/100
The focus axis scores 0 because the snapshot splits the field distribution evenly: 0.5 weight to "robotics / robot learning (machine learning)" and 0.5 to "robot learning and 3D computer vision". Since both labels describe closely overlapping robot-learning research, the zero likely reflects label granularity rather than genuine breadth; both reviewed papers sit squarely in learning-based manipulation.
Collaboration48/100
The collaboration score is 48, based on 10 distinct coauthors in the underlying snapshot. The frequent-collaborator list spans a robotics/robot-learning cluster — Bowen Wen, Stan Birchfield, Yunzhu Li, and Deva Ramanan — alongside earlier-career collaborators including Shahram Najam Syed, Arthur Jakobsson, Prayuj Sachdev, Bardienus P. Duisterhof, Kaifeng Zhang, and Adam Hung. On both 2026 catalog papers Ichnowski is last author (non-corresponding), consistent with a senior advisory role bridging robot learning and 3D vision efforts.
Activity—
Not yet scored: the snapshot lacks a year-by-year publication record, so publishing tempo and consistency cannot be computed. This axis would measure recent output cadence — papers per year in recent windows — and flag gaps or bursts in research activity over time.
Reproducibility78/100
Reproducibility scores 78, computed as the mean over the 2 papers Evolve reviewed (reproMean 78, reproPapers 2). PointZero carries a high readiness rating, while the VLA adaptation study is rated medium, reflecting its proxy-metric and caching caveats. The two-paper base is thin, but both works provide detailed diagnostics — oracle-selection caveats and pre-fine-tuning cost estimates — that support independent verification.
Research summary
Jeffrey Ichnowski is a researcher working at the intersection of robot learning and 3D computer vision; this profile is built from public bibliographic data and Evolve's reviews of two 2026 arXiv papers. On both he appears as last author — a senior, supervisory position — though neither lists him as corresponding author. The first, PointZero, introduces 3D point-track completion as a pre-training objective for learning transferable 3D dynamics, applied to world models, deformable objects, and robot manipulation; Evolve finds the gains large and transfer solid, while noting that headline metrics rely on ground-truth oracle selection. The second diagnoses adaptation in vision-language-action models, showing that parameter-efficient fine-tuning cost is shift-specific and measurable from 10 unlabeled observations before fine-tuning, using LoRA rank allocation and CKA representation similarity. The reproducibility axis scores 78, a mean across the two reviewed papers rated high and medium readiness respectively. Collaboration scores a moderate 48 with 10 distinct coauthors. Output, impact, and activity remain unscored for lack of data, and no institution, country, or ORCID appears in the snapshot.
Collaboration pattern
The collaboration score is 48, with 10 distinct coauthors in the snapshot. Recurring collaborators span a robotics/robot-learning cluster — Bowen Wen, Stan Birchfield, Yunzhu Li, Deva Ramanan — plus earlier-career collaborators including Shahram Najam Syed, Arthur Jakobsson, Prayuj Sachdev, Bardienus P. Duisterhof, Kaifeng Zhang, and Adam Hung. On both 2026 papers Ichnowski is last author but not corresponding, a pattern consistent with senior advisory involvement across robot learning and 3D vision projects.
Research keywords
Highlights
PointZero: point-track pre-training yields large 3D dynamics gains
Diagnosing which VLA layers need tuning — before fine-tuning
Timeline
- 2026Last author on "PointZero: 3D Point Track Completion for Learning Transferable 3D Dynamics" (arXiv), introducing point-track-completion pre-training for world models and robot manipulation.
- 2026Last author on "Not All Layers Need Tuning", showing that vision-language-action fine-tuning cost is shift-specific and predictable from 10 unlabeled observations.
- 2026Both catalog papers land in the same year, consolidating a last-author role spanning robot learning and 3D computer vision.
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.
- Bibliometric scores are computed from public bibliometric snapshots (OpenAlex-style data); no citation or venue-tier metrics were provided.
- Institution, country, and ORCID are absent from the snapshot; the profile is name-resolved from public bibliographic records only.
- Output, impact, and activity axes are unscored because the underlying publication-count and citation data are insufficient.
- The reproducibility score (78) is a mean over only 2 reviewed papers and may not represent the wider research portfolio.