Boston University machine learning researcher spanning LLM post-training with RL and tabular foundation models
3
papers here
3
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
1
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-16) 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—
Total publication counts and venue-level productivity are not yet available for this OpenAlex cluster, so output is not scored. This axis would measure cumulative research production and its distribution across publication venues over time.
Impact—
Citation-based impact measures are thin in this cluster — h-index 1 and i10-index 0 — which is too sparse to score reliably, so impact is left unscored. This axis would measure citation uptake, scholarly influence, and the reach of the work.
Focus0/100
The focus score of 0 reflects a broad footprint: the field distribution splits almost exactly into thirds across LLM post-training with reinforcement learning and tabular deep learning / foundation models, and OpenAlex topics spread across topic modeling and NLP techniques (4 each), reinforcement learning in robotics (3), and neural network applications (2 each). With a sparse 2017-2026 record, this reads as early-stage breadth rather than established concentration.
Collaboration62/100
A collaboration score of 62 rests on 21 distinct coauthors across 8 works in the last three years, roughly 2.6 collaborators per work. All three Evolve-reviewed 2026 papers place the researcher in a middle, non-corresponding position, pointing to team-based machine learning research. Recurring names include Chris Kong, Yefan Tao, Mingyu Chen, Xuezhou Zhang, and Danielle Maddix, suggesting a stable core group spanning both the RL and tabular foundation model lines.
Activity64/100
An activity score of 64 is driven by an accelerating recent cadence: 8 works in the last three years, of which 6 appeared in 2026 and 2 in 2025, against a single earlier record in 2017. Three of the 2026 outputs are arXiv preprints already reviewed by Evolve, indicating an intensive current phase concentrated on LLM post-training and tabular foundation models rather than a long steady publication history.
Reproducibility70/100
Reproducibility averages 70 across the 3 Evolve-reviewed papers. The Mitra-v2 technical report rates high — its evaluation is described as unusually transparent, though the top rank is a statistical tie and v1-to-v2 gains are unattributed. The two RL papers, MInTRL and Cliff, rate medium: MInTRL reports no seed variance, and Cliff is a single run without significance tests and with an unnamed SOTA teacher, so exact replication risk is moderate.
Research summary
Gerald Friedland is a machine learning researcher at Boston University; this profile is name-resolved from public OpenAlex bibliographic data rather than biographical sources. The record is recent and fast-moving: OpenAlex counts 8 works in the last three years, including 6 in 2026 and 2 in 2025, with a single earlier record in 2017. Evolve has reviewed three 2026 arXiv papers, each with the researcher in a middle author position: MInTRL, which injects sparse teacher corrections into on-policy RL rollouts; Cliff, which learns process rewards from a model's first mistake; and the Mitra-v2 technical report on a 77M-parameter tabular foundation model pretrained on synthetic data. The work clusters around LLM post-training with reinforcement learning and tabular deep learning, with the field distribution split roughly into thirds across three subfields — consistent with a focus score of 0. Evolve's reproducibility read across the three papers averages 70, with Mitra-v2 rated high readiness and the two RL papers medium. Citation metrics in this cluster are sparse (h-index 1, i10-index 0), so output and impact are reported as insufficient data rather than scored.
Collaboration pattern
The collaboration footprint is moderately broad: 21 distinct coauthors across 8 works in the last three years, yielding a collaboration score of 62. All three Evolve-reviewed 2026 papers are middle-position, non-corresponding contributions, consistent with team-based research. Recurring coauthors include Chris Kong, Yefan Tao, Mingyu Chen, Xuezhou Zhang, Peixuan Han, and Danielle Maddix, with the same names spanning both the LLM post-training and tabular foundation model lines.
Research keywords
Highlights
Sparse teacher corrections boost on-policy RL (MInTRL)
Transparent evaluation for a 77M tabular foundation model (Mitra-v2)
Process rewards learned from the first mistake (Cliff)
Timeline
- 2017Earliest work recorded for this author cluster in the OpenAlex counts (1 work).
- 20252 works recorded in OpenAlex, marking renewed activity after the sparse earlier record.
- 2026Most productive year on record: 6 works, including three arXiv papers reviewed by Evolve.
- 2026Two distinct lines emerge in the reviewed work: LLM post-training with RL (MInTRL, Cliff) and tabular foundation models (Mitra-v2), all in middle author positions.
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.
- Bibliometrics are computed from OpenAlex only; no ORCID is linked and no other citation source was used.
- Citation counts in this cluster are sparse (h-index 1, i10-index 0), so output and impact are unscored; the low counts may reflect recent author disambiguation rather than career output.
- Gerald Friedland is a relatively common name; this profile is name-resolved from public OpenAlex bibliographic data and the reviewed papers, not from biographical sources.
- All three reviewed papers are 2026 arXiv preprints and may not yet be peer-reviewed.