We thank all reviewers for the constructive comments and are glad to see overall positive reviews toward this work. Response to R1: Thank you for acknowledging contributions of our work and offering helpful suggestions. The symmetric trick is an intuitive way, while our Eq.(1) is more interpretable and can degenerate to the It ignores the directed structure and may lead to degenerated performance. Thus, transition matrix is used to find directly connected neighbors. Thank you for invaluable feedback. We chose a three layer model and set it hidden dimension to 32.

We thank the reviewers for their thorough and positive reviews. We will of course incorporate all the suggested edits by the reviewers as well as more clarifications. We will restate the theorem statement so it would state precisely what is proven. In this paper, we chose to derive the generalization bounds using Graph dimension and VC-dimension. Thank you for your positive review.

With the rapid advancement of AI models, their deployment across diverse tasks has become increasingly widespread. A notable emerging application is leveraging AI models to assist in reviewing scientific papers. However, recent reports have revealed that some papers contain hidden, injected prompts designed to manipulate AI reviewers into providing overly favorable evaluations. In this work, we present an early systematic investigation into this emerging threat. We propose two classes of attacks: (1) static attack, which employs a fixed injection prompt, and (2) iterative attack, which optimizes the injection prompt against a simulated reviewer model to maximize its effectiveness. Both attacks achieve striking performance, frequently inducing full evaluation scores when targeting frontier AI reviewers. Furthermore, we show that these attacks are robust across various settings. To counter this threat, we explore a simple detection-based defense. While it substantially reduces the attack success rate, we demonstrate that an adaptive attacker can partially circumvent this defense. Our findings underscore the need for greater attention and rigorous safeguards against prompt-injection threats in AI-assisted peer review.

We thank all reviewers for the constructive comments and are glad to see overall positive reviews toward this work. Response to R1: Thank you for acknowledging contributions of our work and offering helpful suggestions. The symmetric trick is an intuitive way, while our Eq.(1) is more interpretable and can degenerate to the It ignores the directed structure and may lead to degenerated performance. Thus, transition matrix is used to find directly connected neighbors. Thank you for invaluable feedback. We chose a three layer model and set it hidden dimension to 32.

In this rebuttal, we respond to each reviewer's minor comments individually, as there are no Thank you for the extremely positive review. We will fix the one typo reported. Rubanova et al. [2020] also consider Bayesian optimisation over strings. We will discuss this contemporaneous work (published at the start of this month) in our final version.

We thank the reviewers for their thorough and positive reviews. We will of course incorporate all the suggested edits by the reviewers as well as more clarifications. We will restate the theorem statement so it would state precisely what is proven. In this paper, we chose to derive the generalization bounds using Graph dimension and VC-dimension. Thank you for your positive review.

This research examines whether Airbnb guests' positive and negative comments influence acceptance rates and rental prices across six U.S. regions: Rhode Island, Broward County, Chicago, Dallas, San Diego, and Boston. Thousands of reviews were collected and analyzed using Natural Language Processing (NLP) to classify sentiments as positive or negative, followed by statistical testing (t-tests and basic correlations) on the average scores. The findings reveal that over 90 percent of reviews in each region are positive, indicating that having additional reviews does not significantly enhance prices. However, listings with predominantly positive feedback exhibit slightly higher acceptance rates, suggesting that sentiment polarity, rather than the sheer volume of reviews, is a more critical factor for host success. Additionally, budget listings often gather extensive reviews while maintaining competitive pricing, whereas premium listings sustain higher prices with fewer but highly positive reviews. These results underscore the importance of sentiment quality over quantity in shaping guest behavior and pricing strategies in an overwhelmingly positive review environment.

Deceptive reviews are becoming increasingly common, especially given the increase in performance and the prevalence of LLMs. While work to date has addressed the development of models to differentiate between truthful and deceptive human reviews, much less is known about the distinction between real reviews and AI-authored fake reviews. Moreover, most of the research so far has focused primarily on English, with very little work dedicated to other languages. In this paper, we compile and make publicly available the MAiDE-up dataset, consisting of 10,000 real and 10,000 AI-generated fake hotel reviews, balanced across ten languages. Using this dataset, we conduct extensive linguistic analyses to (1) compare the AI fake hotel reviews to real hotel reviews, and (2) identify the factors that influence the deception detection model performance. We explore the effectiveness of several models for deception detection in hotel reviews across three main dimensions: sentiment, location, and language. We find that these dimensions influence how well we can detect AI-generated fake reviews.