AI Paper Writing and Reviewing: Empirical Workshop Observations
TuhinChakr · x · 2026-07-16
This post summarizes an analysis of AI-generated content at the Mech Interp Workshop, with the following key findings:
- A significant increase in submissions, particularly from single-author and repeat first-author papers
- A higher proportion of these submissions are flagged as AI-written
- High-quality papers remain predominantly human-written
- However, AI reviewers are more likely than human reviewers to give higher scores to AI-written papers
- Human reviewers generally find these submissions difficult to read, sometimes needing to re-read the abstract multiple times while still finding the contribution unclear
The value of this post lies in its breakdown of how AI-generated content impacts academic workflows into observable phenomena: shifts in submission demographics, reviewing biases, and inconsistencies between human and AI reviewing standards.
Related event: Mech Interp Workshop Sees Surge in AI-Generated Submissions(2 posts)→
More from Research
- Chinese AI labs are now treating distillation obfuscation as the top research topic — pmddomingos · 2026-07-22
- Structural ensembles beat single predictions in TCR:pMHC generalization study — quaidmorris · 2026-07-22
- RSS launches under OMSF to push structural biology data modeling at scale — MoAlQuraishi · 2026-07-22
- enFoldX tops 8 neoantigen scans and an unseen-peptide benchmark — quaidmorris · 2026-07-22
- enFoldX reaches AUC 0.82 on human VDJdb and transfers to mouse at 0.76 — quaidmorris · 2026-07-22
- enFoldX gains accuracy as AF3 ensemble disagreement rises for non-binders — quaidmorris · 2026-07-22