Gaze Patterns Predict Preference and Confidence in Pairwise AI Image Evaluation
Nikolas Papadopoulos, Shreenithi Navaneethan, Sheng Bai, Ankur Samanta, Paul Sajda · 2026
Read Gaze Patterns Predict Preference and Confidence in Pairwise AI Image Evaluation by Nikolas Papadopoulos, Shreenithi Navaneethan, Sheng Bai, Ankur Samanta, Paul Sajda (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-24849.txt.
Description
Papadopoulos and co-authors use eye-tracking to study preference in pairwise AI image choice. Preference learning relies on human pairwise judgments, yet the process behind them is little understood. Thirty participants did 1,800 trials while their gaze was recorded. The gaze cascade effect replicates: gaze shifts toward the chosen image about one second before the decision. Gaze features predict the binary choice with 68 percent accuracy. Transitions between images also separate confident from uncertain decisions, with low-confidence trials showing more switches per second.
Catalog record
| CREATOR | Nikolas Papadopoulos, Shreenithi Navaneethan, Sheng Bai, Ankur Samanta, Paul Sajda |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25518 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.24849