How Sampling Affects the Detectability of Machine-written texts: A Comprehensive Study
Bibliographic record
Abstract
As texts generated by Large Language Models (LLMs) are ever more common and often indistinguishable from human-written content, research on automatic text detection has attracted growing attention.Many recent detectors report near-perfect accuracy, often boasting AUROC scores above 99%.However, these claims typically assume fixed generation settings, leaving open the question of how robust such systems are to changes in decoding strategies.In this work, we systematically examine how sampling-based decoding impacts detectability, with a focus on how subtle variations in a model's (sub)word-level distribution affect detection performance.We find that even minor adjustments to decoding parameters -such as temperature, top-p, or nucleus sampling -can severely impair detector accuracy, with AUROC dropping from near-perfect levels to 1% in some settings.Our findings expose critical blind spots in current detection methods and emphasize the need for more comprehensive evaluation protocols.To facilitate future research, we release a large-scale dataset encompassing 37 decoding configurations, along with our code and evaluation framework https://github.com/ BaggerOfWords/Sampling-and-Detection.t y p ic a l: 0 .3t y p ic a l: 0 .5 t y p ic a l: 0 .7 t y p ic a l: 0 .8t y p ic a l: 0 .9t y p ic a l: 0 .9 5 e t a : 1 e -4 e t a : 1 e -3 e t a : 5 e -3 e t a : 0 .0 1 e t a : 0 .0 5 e t a : 0 .1 m ix t u r e : 4 7 .0temp: 0.5 temp: 0.7 temp: 0.9 temp: 1.0
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 | 0.012 | 0.095 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.003 | 0.005 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.002 | 0.003 |
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".