Alibaba Group researcher on LLM agent reinforcement learning: credit assignment and rubric-grounded self-evolution
2
papers here
1
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-17) 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: this snapshot contains no verified publication counts, venue history, or yearly output volume for this name. The axis would measure cumulative research output across venues and years.
Impact—
Not yet scored: no citation counts, h-index, or downstream-adoption signals are available in this snapshot. The axis would measure how widely and influentially the work is cited and used by others.
Focus100/100
Focus scores 100/100: the field distribution derived from the reviewed work maps entirely to a single label, machine learning (LLM agents and reinforcement learning). Both 2026 papers, ArenaFlow and ARISE-RL, sit inside that one area, working on open-ended agent RL training, credit assignment, and reward design. This is a maximally concentrated specialist profile, though it is computed from only two reviewed papers.
Collaboration62/100
Collaboration scores 62/100, drawn from 22 distinct coauthors observed around the reviewed work. Frequently recurring names include Ruixue Ding, Shihang Wang, Pengjun Xie, Zheng-Jun Zha, Xi Chen, Qiang Zhang, Fanrui Zhang, Yinfeng Huang, Yi Zheng, Kaipeng Zhang, and Jiawei Liu. On both 2026 papers Boli Chen holds a middle author position without corresponding-author role, indicating a contributor role within larger research teams.
Activity—
Not yet scored: the snapshot lacks multi-year publication frequency data, so recent research cadence cannot be computed. The axis would measure output regularity and research engagement over recent years.
Reproducibility55/100
Reproducibility scores 55/100, a mean computed over a single reviewed paper (reproPapers = 1, reproMean = 55). Both cataloged papers carry medium reproducibility readiness. Evolve's verdicts name concrete gaps: ArenaFlow shows judge circularity and lacks variance tests, while ARISE-RL evaluates on self-built benchmarks, so reported gains are strong but not yet independently verifiable.
Research summary
Boli Chen is a researcher at Alibaba Group (China) whose cataloged work centers on machine learning for LLM agents and reinforcement learning. The Evolve catalog holds two 2026 arXiv papers, both with Chen in a middle author position and neither as corresponding author. ArenaFlow extends tournament-based agent RL with step- and skill-level credit propagation; Evolve records a +19.5 improvement over ArenaRL on Open-Travel while flagging judge circularity and missing variance tests. ARISE-RL claims state-of-the-art open-ended agent RL through rubric-grounded generator-solver co-evolution and reward-gated self-distillation, with gains Evolve judges strong but evaluated on self-built benchmarks. The radar reflects this concentration: focus scores 100/100, with the field distribution mapped entirely to one label, and collaboration scores 62/100 across 22 distinct coauthors including Ruixue Ding, Shihang Wang, Pengjun Xie, Zheng-Jun Zha, and Kaipeng Zhang. Reproducibility averages 55/100 over one scored paper, and both papers carry medium reproducibility readiness. Output, impact, and activity dimensions lack sufficient data and are not scored. As Boli Chen is a common name, this profile is name-resolved from public bibliographic data.
Collaboration pattern
The data shows 22 distinct coauthors around Boli Chen's reviewed work. Frequently recurring names include Ruixue Ding, Shihang Wang, Pengjun Xie, Zheng-Jun Zha, Xi Chen, Qiang Zhang, Fanrui Zhang, Yinfeng Huang, Yi Zheng, Kaipeng Zhang, and Jiawei Liu. On both 2026 papers Boli Chen sits in a middle author position without a corresponding-author role, consistent with contributing within larger reinforcement-learning teams at Alibaba Group rather than leading small groups.
Research keywords
Highlights
ArenaFlow: hierarchical credit propagation for open-ended agent RL
ARISE-RL: rubric-grounded self-evolution for agent RL
Timeline
- 2026ArenaFlow appears on arXiv, extending tournament-based agent RL with step- and skill-level credit propagation; Evolve records a +19.5 gain over ArenaRL on Open-Travel.
- 2026ARISE-RL appears on arXiv, proposing rubric-grounded generator-solver co-evolution with reward-gated self-distillation for open-ended agent RL.
- 2026Both papers enter the Evolve catalog in the same year with medium reproducibility readiness, forming a concentrated open-ended agent RL profile at Alibaba Group.
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 rest on a thin snapshot: output, impact, and activity are unscored for lack of publication-volume and citation data; reproducibility averages 55 over a single scored paper.
- Boli Chen is a common name; this profile is name-resolved from public bibliographic data and may merge or split same-name researchers.
- Both cataloged papers are 2026 arXiv preprints with Boli Chen in middle-author, non-corresponding positions; no venue history, awards, or career facts are in scope.
- Reproducibility readiness for both papers is rated medium; Evolve flags judge circularity and missing variance tests (ArenaFlow) and self-built benchmarks (ARISE-RL).