Hao Wang: ML researcher across graph domain adaptation and embodied AI task planning; two 2026 arXiv papers.
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-08) 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—
No publication-volume data is attached to this record, so output is unscored. This axis would measure paper count, venue spread, and authorship roles over time — none of it derivable from two reviewed arXiv entries alone.
Impact—
No citation counts, h-index, or influence signals are available in this snapshot, so impact is unscored. This axis would measure citations to the reviewed papers and their uptake in the graph domain adaptation and embodied AI communities.
Focus0/100
Focus scores 0, reflecting an even split of the cataloged field distribution — 0.5 embodied AI / robot learning and 0.5 machine learning (graph representation learning and domain adaptation). In practice the reviewed 2026 papers sit in two unrelated areas: one on graph domain adaptation (EviGDA) and one on event-structured robot task planning (X-Planner). The record therefore shows breadth across fields rather than concentration in a single line of work.
Collaboration73/100
Collaboration scores 73, grounded in 38 distinct coauthors across the record. Twelve recur as frequent collaborators (including Howard Lu, Shalfun Li, Porter Pan, Eric Hu, Robert Wang, and KZ Zheng), suggesting stable team-based projects alongside a wide collaboration surface. Both reviewed papers are middle-authorship team efforts with no corresponding-author role in the catalog.
Activity—
No per-year publication cadence is available in this snapshot, so activity is unscored. This axis would measure publishing rhythm and continuity over recent years; with only two 2026 papers on file, no trend can be established.
Reproducibility38/100
Reproducibility scores 38, the mean over two reviewed papers. Evolve rates the graph domain adaptation paper (EviGDA) as medium repro-readiness, while X-Planner is rated low: the review notes no ablations, no significance testing, and offline results trailing kimi3, which limits independent verification of the claimed real-robot gains.
Research summary
Hao Wang is an early-stage machine learning researcher whose cataloged work spans two distinct areas: graph representation learning and domain adaptation on one side, and embodied AI / robot learning on the other. The Evolve snapshot resolves two 2026 arXiv papers, in both of which Hao Wang is a middle-position author and not the corresponding author. The first, 'Graph Domain Adaptation Does Not End with Representation Learning,' presents EviGDA, which claims state-of-the-art graph domain adaptation by fusing an adapted graph expert with a source-only MLP; Evolve's review finds the results solid but characterizes the contribution as largely principled ensembling rather than new adaptation machinery. The second, 'X-Planner: Event-Structured Task Planning for Embodied Intelligence,' pairs event-structured planning data with parallel latent chain-of-thought reasoning for robot manipulation; Evolve notes large real-robot wins but flags missing ablations and significance tests, with offline performance trailing kimi3. The field distribution splits evenly (0.5 and 0.5) between the two areas, producing a focus score of 0, while 38 distinct coauthors yield a collaboration score of 73. Reproducibility scores 38, the mean over two reviewed papers rated medium and low respectively. Output, impact, and activity lack underlying data and are left unscored.
Collaboration pattern
With 38 distinct coauthors and a collaboration score of 73, Hao Wang's cataloged work is team-based and broad. Twelve collaborators recur frequently — including Howard Lu, Shalfun Li, Porter Pan, Eric Hu, Lily Li, Maeve Zhang, Robert Wang, KZ Zheng, and Viggo Chen — pointing to a stable core group plus a rotating wider periphery. On both reviewed 2026 papers Hao Wang is a middle author, with no corresponding-author role in the catalog.
Research keywords
Highlights
EviGDA: pushing graph domain adaptation past representation learning
X-Planner: event-structured planning with parallel latent chain-of-thought
Timeline
- 2026Co-authored 'Graph Domain Adaptation Does Not End with Representation Learning' (arXiv); EviGDA claims graph domain adaptation state of the art by fusing an adapted graph expert with a source-only MLP.
- 2026Co-authored 'X-Planner: Event-Structured Task Planning for Embodied Intelligence' (arXiv), a middle-authorship contribution to event-structured planning data and parallel latent chain-of-thought for robot manipulation.
- 2026Catalog snapshot shows the year's output split evenly across two fields (0.5 embodied AI / robot learning, 0.5 graph machine learning and domain adaptation), with 38 distinct coauthors on record.
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.
- Name only: no institution, country, or ORCID is attached to this record; 'Hao Wang' is a common name, so this profile is name-resolved from public bibliographic data and may merge or split identities.
- Radar scores come from Evolve's bibliometric snapshot; output, impact, and activity lack underlying data and are left unscored rather than estimated.
- Scope is limited to two reviewed arXiv papers (both 2026); venue publications, earlier work, and citation counts are not represented.
- The reproducibility score (38) is a mean over only 2 reviewed papers, so it is a coarse signal.