Iclr Points

iclr-points

ICLR Points: How Many ICLR Publications Is One Paper in Each Area? — Luo et al. (2025) (arXiv:2503.16623, 2025)

What this evaluates

Quantifies the average research effort required to produce one publication at top-tier conferences across 27 computer science subfields. It enables cross-area comparisons of faculty productivity and publication effort by normalizing faculty headcounts against publication counts.

Datasets

• DBLP publication data (2019–2023) and faculty headcounts — total ?; splits: all (-1)

Metrics

• ICLR points (primary) — range: other
  • ICLR points = (Total faculty count in a subfield) / (Total publications in that subfield's top-tier conferences). It represents the average faculty effort required per publication.

Input / output format

Input: Faculty headcount and publication counts per computer science subfield from DBLP (2019–2023).

Output: Scalar value representing average research effort per publication in ICLR points for each of the 27 subfields.

Scoring recipe

def compute_iclr_points(faculty_count, publication_count):
    if publication_count == 0:
        return 0
    return faculty_count / publication_count

Common pitfalls

• Assumes faculty headcount is a direct proxy for research effort, ignoring part-time roles or varying workloads.
• Does not reflect publication quality, citation impact, or inherent field difficulty.
• Limited by the assumption that faculty publications dominate the area's output.

Evidence (verbatim from paper)

Based on the datasets and methodology described above, we compute the average effort to produce one publication in each area's top-tier conferences in the unit of ICLR points.

Citation

@misc{luo2025iclrpoints,
  title={ICLR Points: How Many ICLR Publications Is One Paper in Each Area?},
  author={Luo et al. (2025)},
  year={2025},
  note={arXiv:2503.16623}
}

• arXiv: 2503.16623