Bibliographic record
Abstract
Following the theme of my August and September editorials on peer review, this month I will reflect on "Peer Review Week 2023" (held September 25-29) and the use of large language models (LLMs) in peer review.Started in 2016, Peer Review Week is a global event that recognizes and celebrates the importance of peer review to research quality.Each year, a focus topic is selected, and a number of organizations develop content designed to address the primary topic.For 2023, the topic was "peer review and the future of publishing," ". . . to highlight the changing publishing landscape and the ongoing vital role of peer review in shaping scholarly communication."The problem of finding reviewers is one that I have addressed in those previous editorials, spanning across all publishing disciplines from medical research to our field of food science.Some webinar titles included "The Future of Peer Review," "The Great Peer Review Debate: Open, Closed and Transparent Models," "Best Practices in Peer Review," "The Future of Peer Review: Diversification and Decentralization," and "Training Peer Reviewers as a Form of Engagement: A Solution to the Peer Review Crisis."These webinars were put on by a variety of diverse organizations, from individual publishers to organizations like the Council of Scientific Editors (CSE).You can view this list of events, including recordings of many, at https://peerreviewweek.wordpress.com/eventsand-activities-2023/.Amanda Ferguson, Director of IFT Scientific Journals, attended two webinars.First, Canadian Science and Medical Editors Network's webinar on "Training Peer Reviewers as a Form of Engagement: A Solution to the Peer Review Crisis?" emphasized the need for journals to foster a sense of community among their editors and reviewers to increase reviewer engagement.The speakers discussed how many reviewers do not understand what their role is supposed to be: to validate, critique, and help research
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.028 | 0.104 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.004 | 0.002 |
| Bibliometrics | 0.005 | 0.003 |
| Science and technology studies | 0.008 | 0.004 |
| Scholarly communication | 0.036 | 0.012 |
| Open science | 0.004 | 0.012 |
| Research integrity | 0.011 | 0.012 |
| Insufficient payload (model declined to judge) | 0.410 | 0.585 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".