Large language models (LLMs) are increasingly used to generate, classify, and annotate data whose outputs feed downstream hypothesis tests. However, LLM-based research is easy to p-hack: a researcher can tune the prompts, decoding parameters, or output format until a desired result is reached. We propose a protocol to mitigate p-hacking in LLM-based research: preregistering the experiment and eligible models, and then running it on the first eligible LLM that is released after the preregistration. The researcher finalizes the procedure on current models, preregisters the analysis plan together with a set of eligible future models, and runs the confirmatory analysis on the first eligible model released afterward. Because this model does not exist at commitment time, it cannot be hacked against; furthermore, configurations that hack one model frequently do not transfer to the next. We evaluate the protocol on two tasks whose true values are known. Across 20 models from four providers and 11 LLM-analysis configurations, the protocol would have blocked successful transfer of the p-hack in 73.9

Competitive selection processes, from scientific funding to admissions and hiring, use evaluations to score candidates, and eventually choose a subset of them based on those scores. Recently, many organizations have adopted partial lotteries, which randomize selection based on evaluation scores. However, existing lottery designs are inherently unstable, as a small change to a single candidate's score can cause large shifts in their selection probabilities. This instability undermines a key goal of lotteries: reducing the influence of fine-grained score distinctions near the decision boundary. We propose smoothness as a design principle for partial lotteries, formalizing it as a Lipschitz condition on the mapping from review scores over candidates to selection probabilities. We introduce the Clipped Linear Lottery, a simple mechanism in which selection probabilities scale linearly with estimated quality between an upper threshold, above which we always accept, and a lower threshold, below which we always reject. We prove that the Clipped Linear Lottery's worst-case regret matches a lower bound for any smooth selection rule up to a factor of (1 - k/n), where k/n is the acceptance rate. We compare smooth selection to other stability notions like Individual Fairness and Differential Privacy, showing that the Clipped Linear Lottery achieves a better smoothness-regret tradeoff than alternatives. Experiments on real peer review data from ICLR 2025, NeurIPS 2024, and the Swiss National Science Foundation demonstrate that existing lottery designs are highly unstable in practice even under perturbations to a single score. Our experiments also confirm the tightness of our theoretical analysis and show that our proposed Clipped Linear Lottery achieves a better smoothness-utility tradeoff than alternatives in practice.

A number of scientific conferences and journals have recently enacted policies that prohibit LLM usage by peer reviewers, except for polishing, paraphrasing, and grammar correction of otherwise human-written reviews. But, are these policies enforceable? To answer this question, we assemble a dataset of peer reviews simulating multiple levels of human-AI collaboration, and evaluate five state-of-the-art detectors, including two commercial systems. Our analysis shows that all detectors misclassify a non-trivial fraction of LLM-polished reviews as AI-generated, thereby risking false accusations of academic misconduct. We further investigate whether peer-review-specific signals, including access to the paper manuscript and the constrained domain of scientific writing, can be leveraged to improve detection. While incorporating such signals yields measurable gains in some settings, we identify limitations in each approach and find that none meets the accuracy standards required for identifying AI use in peer reviews. Importantly, our results suggest that recent public estimates of AI use in peer reviews through the use of AI-text detectors should be interpreted with caution, as current detectors misclassify mixed reviews (collaborative human-AI outputs) as fully AI generated, potentially overstating the extent of policy violations.

In this position paper, we enumerate a number of problems with the current peer-review process based on extensive empirical evidence. We argue for two structural reforms: (1) separating publication from presentation via a four-step process that first evaluates correctness, publishes all sound papers, then uses community-based ratings to select presentations; and (2) offering parallel anonymous and non-anonymous review tracks, where the non-anonymous track releases all review data publicly to increase accountability and generate valuable research datasets. We argue how our proposed policies can mitigate these problems. We urge the community to leverage the learnings from the experiments conducted in peer-review processes and incorporate evidence-based policy design.

There is growing interest in applying artificial intelligence (AI) to automate and support complex decision-making tasks. However, it remains unclear how algorithms compare to human judgment in contexts requiring semantic understanding and domain expertise. We examine this in the context of the judge assignment problem, matching submissions to suitably qualified judges. Specifically, we tackled this problem at the Harvard President’s Innovation Challenge, the university’s premier venture competition awarding over $500,000 to student and alumni startups. This setting represents a real-world environment where high-quality judge assignment is essential. We developed an AI-based judge assignment algorithm, the Hybrid Lexical-Semantic Similarity Ensemble (HLSE), and deployed it at the competition. We then evaluated its performance against human expert assignments using blinded match-quality scores from judges on 309 judge-venture pairs. Using a Mann-Whitney U statistic-based test, we found no statistically significant difference in assignment quality between the two approaches (AUC=0.48, p=0.40); on average, algorithmic matches were rated 3.90 and manual matches 3.94 on a 5-point scale, where 5 indicates an excellent match. Furthermore, manual assignments that previously required a full week could be automated in several hours by the algorithm during deployment. These results demonstrate that HLSE achieves human-expert-level matching quality while offering greater scalability and efficiency, underscoring the potential of AI-driven solutions to support and enhance human decision-making for judge assignment in high-stakes settings.

Research evaluation systems – including journals, conferences, and funders – are increasingly using author-level submission limits to manage growing submission loads. Most existing policies charge each submission as a unit cost against every coauthor's quota. This treats a solo-authored submission and a large collaborative submission identically for each author, even though the reviewing demand of a collaborative submission is jointly attributable to many authors rather than one. Thus we ask the question: how many submissions may an author make under coauthorships? We propose a "Harmonic Quota Rule", in which an author's cost for a submission decreases with the number of coauthors as the reciprocal of their harmonic number. We derive this rule in a principled manner that navigates the tension between respecting collaborations and being resistant to manipulation by adding spurious authors. We also develop a Generalized Harmonic Quota Rule, a framework that subsumes the Harmonic Quota Rule and other natural quota rules. Our framework requires specification of only three interpretable parameters, thereby enabling organizers to choose among various seemingly disparate rules. Our work may also be useful in other scarce-resource allocation settings, such as allocation of compute and telescope time. An interactive tool is available at https://www.cs.cmu.edu/ nihars/quota/organizer.html

Scientific fraud is the instrument of doubt that malicious entities can use to establish controversy in science. Historically, it required the resources of a company: deep pockets, ghostwritten articles, and corrupt academics. Today, Artificial Intelligence (AI) is increasingly automating scientific research, so we ask: Can a remote adversary weaponize the honest use of AI in science to compromise scientific integrity? We envision and empirically evaluate a new attack, indirect data poisoning, in which an adversary corrupts an open dataset and uploads the poisoned variant to a public repository. Autonomous research agents may independently retrieve and process this data, turning honest scientists into the unpaid and unwitting distributors of fraud at scale. Across five socially-salient topics, from hiring discrimination to the safety of autonomous vehicles, three widely used frontier AI systems (Claude Code with Claude Opus 4.7, Codex with GPT-5.5, Gemini CLI with Gemini 3.1 Pro), and 450 ethically contained experimental runs, we find that poisoning succeeds in 49.56

In many applications, human and LLM evaluators use assessments of relevant criteria to create an overall evaluation for an item or individual. For example, in admissions, committees assess candidates on attributes such as test scores, GPA, and research experience to evaluate their overall fit for the program. Another example arises in medical care where clinicians use patient reports of symptoms to consider preliminary diagnoses and assess risks. Each setting involves mapping multiple criteria to an overall evaluation – a process that reflects the evaluator's underlying preferences. We focus on the fundamental question of learning these preferences. Many applications of this problem make specific modeling assumptions on evaluator preferences that may be substantially violated in the real world. We make the minimal assumption that the preference function is coordinate-wise non-decreasing, which is reasonable in a large number of evaluation settings. We theoretically characterize the severity of model mismatch for many common assumptions and show that it can lead to significant issues for learning evaluator preferences and other important downstream tasks. We then present an algorithm for learning evaluators' preferences that is robust to model mismatch. We prove theoretically that our algorithm can learn any preference function without sacrificing performance when the linearity assumption holds. Evaluations of our algorithm with synthetic simulations and real-world data confirm its ability to learn preferences robustly and illustrate key aspects of LLM and human preferences.

Reproducing research results from papers and released code is central to scientific progress. Existing works have introduced benchmarks to evaluate whether LLM agents can assist with reproducibility, but they are difficult to scale due to their reliance on substantial manual effort for data curation and evaluation. We introduce ReproRepo, a scalable framework for reproducibility evaluation that leverages human-raised GitHub issues as naturally occurring supervision on realistic reproduction blockers. We instantiate ReproRepo on 1,149 recent machine learning papers from major conferences and evaluate four frontier model-agent configurations. Our results show that LLM agents, even without executing code, can identify many real-world reproducibility problems from paper-repository pairs: the best agent in our study, namely Codex with GPT-5.5, surfaces at least one semantically related human-reported blocker for 90

Many decision-making processes involve evaluating and selecting items, including scientific peer review, job hiring, school admissions, and investment decisions. These domains feature error-prone evaluations and uncertainty about outcomes, which undermine deterministic selection rules. Consequently, randomized selection mechanisms are gaining traction. However, current randomized approaches are ad hoc and, as we prove, inappropriate for their purported objectives. We propose a principled framework for randomized decision-making based on interval estimates of item quality. We introduce MERIT (Maximin Efficient Randomized Interval Top-$k$), which maximizes the worst-case expected number of top candidates selected under uncertainty represented by overlapping intervals. MERIT provides optimal resource allocation under an interpretable robustness notion. We develop a polynomial-time, practically efficient algorithm and prove our approach satisfies desirable axiomatic properties not guaranteed by existing methods. Experiments on synthetic peer review data from grant funding and conferences demonstrate that MERIT matches existing algorithms' expected utility under fully probabilistic models while outperforming them under our worst-case formulation.

Is it possible to reliably evaluate the quality of peer reviews? We study this question driven by two primary motivations - incentivizing high-quality reviewing using assessed quality of reviews and measuring changes to review quality in experiments. We conduct a large scale study at the NeurIPS 2022 conference, a top-tier conference in machine learning, in which we invited (meta)-reviewers and authors to voluntarily evaluate reviews given to submitted papers. First, we conduct a randomized controlled trial to examine bias due to the length of reviews. We generate elongated versions of reviews by adding substantial amounts of non-informative content. Participants in the control group evaluate the original reviews, whereas participants in the experimental group evaluate the artificially lengthened versions. We find that lengthened reviews are scored (statistically significantly) higher quality than the original reviews. Additionally, in analysis of observational data we find that authors are positively biased towards reviews recommending acceptance of their own papers, even after controlling for confounders of review length, quality, and different numbers of papers per author. We also measure disagreement rates between multiple evaluations of the same review of 28% - 32%, which is comparable to that of paper reviewers at NeurIPS. Further, we assess the amount of miscalibration of evaluators of reviews using a linear model of quality scores and find that it is similar to estimates of miscalibration of paper reviewers at NeurIPS. Finally, we estimate the amount of variability in subjective opinions around how to map individual criteria to overall scores of review quality and find that it is roughly the same as that in the review of papers. Our results suggest that the various problems that exist in reviews of papers - inconsistency, bias towards irrelevant factors, miscalibration, subjectivity - also arise in reviewing of reviews.

Given a set of items and a set of evaluators who all individually rank them, how do we aggregate these evaluations into a single societal ranking? Work in social choice and statistics has produced many aggregation methods for this problem, each with its desirable properties, but also with its limitations. Further, existing impossibility results rule out designing a single method that achieves every property of interest. Faced with this trade-off between incompatible desiderata, how do we decide which aggregation rule to use, i.e., what is a good rule picking rule? In this paper, we formally address this question by introducing a novel framework for rule picking rules (RPRs). We then design a data-driven RPR that identifies the best aggregation method for each specific setting, without assuming any generative model. The principle behind our RPR is to pick the rule which maximizes the consistency of the output ranking if the data collection process were repeated. We introduce several consistency-related axioms for RPRs and show that our method satisfies them, including those failed by a wide class of natural RPRs. While we prove that the algorithmic problem of maximizing consistency is computationally hard, we provide a sampling-based implementation of our RPR that is efficient in practice. We run this implementation on known statistical models and find that, when possible, our method selects the maximum likelihood estimator of the data. Finally, we show that our RPR can be used in many real-world settings to gain insights about how the rule currently being used can be modified or replaced to substantially improve the consistency of the process. Taken together, our work bridges an important gap between the axiomatic and statistical approaches to rank aggregation, laying a robust theoretical and computational foundation for principled rule picking.

Peer-review venues have increasingly adopted open reviewing policies that publicly release anonymized reviews and permit public commenting. Venues have adopted a variety of policies, and there is still ongoing debate about the benefits and drawbacks of decisions. To inform this debate, we surveyed 2,385 reviewers, authors, and other peer-review participants in machine learning to understand their experiences and opinions. Our key findings are: (a) Preferences: Over 80 (b) Benefits: Respondents cite improved public understanding (75.3 (c) Challenges: The top concern is resubmission bias, where rejection history biases future reviewers (ranked top impact of open reviewing by 41 (d) AI and open peer review: Participants believe open policies deter "AI slop" submissions (71.9 Finally, we use AI to annotate 4,244 reviews from ICLR (fully open) and NeurIPS (partially open). We find that the fully open venue (ICLR) has higher levels of correctness and completeness than the partially open venue (NeurIPS). The effect size is small for correctness and very small for completeness, and both are statistically significant. We also find that there is no statistically significant difference in the level of substantiation. We release the full dataset at https://github.com/justinpayan/OpenReviewAnalysis.

The increasing volume of papers and proposals that undergo peer review emphasizes the pressing need for greater automation to effectively manage the growing scale. In this study, we present the deployment and evaluation of machine learning and optimization techniques to assign proposals to reviewers that were developed for the Atacama Large Millimeter/submillimeter Array (ALMA) during the Cycle 10 Call for Proposals issued in 2023. Using topic modeling algorithms, we identify the proposal topics and assess reviewers’ expertise based on their previous ALMA proposal submissions. We then apply an adapted version of the assignment optimization algorithm from PeerReview4All to maximize the alignment between proposal topics and reviewer expertise. Our evaluation shows a significant improvement in matching reviewer expertise: the median similarity score between the proposal topic and reviewer expertise increased by 51 percentage points compared to the previous cycle, and the percentage of reviewers reporting expertise in their assigned proposals rose by 20 percentage points. Furthermore, the assignment process proved highly effective in that no proposals required reassignment due to significant mismatches, resulting in a savings of 3–5 days of manual effort.

The integrity of peer review is fundamental to scientific progress, but the rise of large language models (LLMs) has introduced concerns that some reviewers may rely on these tools to generate reviews rather than writing them independently. Although some venues have banned LLM-assisted reviewing, enforcement remains difficult as existing detection tools cannot reliably distinguish between fully generated reviews and those merely polished with AI assistance. In this work, we address the challenge of detecting LLM-generated reviews. We consider the approach of performing indirect prompt injection via the paper's PDF, prompting the LLM to embed a covert watermark in the generated review, and subsequently testing for presence of the watermark in the review. We identify and address several pitfalls in naïve implementations of this approach. Our primary contribution is a rigorous watermarking and detection framework that offers strong statistical guarantees. Specifically, we introduce watermarking schemes and hypothesis tests that control the family-wise error rate across multiple reviews, achieving higher statistical power than standard corrections such as Bonferroni, while making no assumptions about the nature of human-written reviews. We explore multiple indirect prompt injection strategies-including font-based embedding and obfuscated prompts-and evaluate their effectiveness under various reviewer defense scenarios. Our experiments find high success rates in watermark embedding across various LLMs. We also empirically find that our approach is resilient to common reviewer defenses, and that the bounds on error rates in our statistical tests hold in practice. In contrast, we find that Bonferroni-style corrections are too conservative to be useful in this setting.

Many decision-making processes involve evaluating and then selecting items; examples include scientific peer review, job hiring, school admissions, and investment decisions. The eventual selection is performed by applying rules or deliberations to the raw evaluations, and then deterministically selecting the items deemed to be the best. These domains feature error-prone evaluations and uncertainty about future outcomes, which undermine the reliability of such deterministic selection rules. As a result, selection mechanisms involving explicit randomization that incorporate the uncertainty are gaining traction in practice. However, current randomization approaches are ad hoc, and as we prove, inappropriate for their purported objectives. In this paper, we propose a principled framework for randomized decision-making based on interval estimates of the quality of each item. We introduce MERIT (Maximin Efficient Randomized Interval Top-k), an optimization-based method that maximizes the worst-case expected number of top candidates selected, under uncertainty represented by overlapping intervals (e.g., confidence intervals or min-max intervals). MERIT provides an optimal resource allocation scheme under an interpretable notion of robustness. We develop a polynomial-time algorithm to solve the optimization problem and demonstrate empirically that the method scales to over 10,000 items. We prove that MERIT satisfies desirable axiomatic properties not guaranteed by existing approaches. Finally, we empirically compare algorithms on synthetic peer review data. Our experiments demonstrate that MERIT matches the performance of existing algorithms in expected utility under fully probabilistic review data models used in previous work, while outperforming previous methods with respect to our novel worst-case formulation.

We introduce the novel problem of benchmarking fraud detectors on private graph-structured data. Currently, many types of fraud are managed in part by automated detection algorithms that operate over graphs. We consider the scenario where a data holder wishes to outsource development of fraud detectors to third parties (e.g., vendors or researchers). The third parties submit their fraud detectors to the data holder, who evaluates these algorithms on a private dataset and then publicly communicates the results. We propose a realistic privacy attack on this system that allows an adversary to de-anonymize individuals' data based only on the evaluation results. In simulations of a privacy-sensitive benchmark for facial recognition algorithms by the National Institute of Standards and Technology (NIST), our attack achieves near perfect accuracy in identifying whether individuals' data is present in a private dataset, with a True Positive Rate of 0.98 at a False Positive Rate of 0.00. We then study how to benchmark algorithms while satisfying a formal differential privacy (DP) guarantee. We empirically evaluate two classes of solutions: subsample-and-aggregate and DP synthetic graph data. We demonstrate through extensive experiments that current approaches fail to provide utility when guaranteeing DP. Our results indicate that the error arising from DP trades off between bias from distorting graph structure and variance from adding random noise. Current methods lie on different points along this bias-variance trade-off, but more complex methods tend to require high-variance noise addition, undermining utility.

Highlighting the critical need for stronger safeguards against identity theft in peer review and academia at large and proposing mitigating strategies.

In the peer review process of top-tier machine learning (ML) and artificial intelligence (AI) conferences, reviewers are assigned to papers through automated methods. These assignment algorithms consider two main factors: (1) reviewers' expressed interests indicated by their bids for papers, and (2) reviewers' domain expertise inferred from the similarity between the text of their previously published papers and the submitted manuscripts. A significant challenge these conferences face is the existence of collusion rings, where groups of researchers manipulate the assignment process to review each other's papers, providing positive evaluations regardless of their actual quality. Most efforts to combat collusion rings have focused on preventing bid manipulation, under the assumption that the text similarity component is secure. In this paper, we demonstrate that even in the absence of bidding, colluding reviewers and authors can exploit the machine learning based text-matching component of reviewer assignment used at top ML/AI venues to get assigned their target paper. We also highlight specific vulnerabilities within this system and offer suggestions to enhance its robustness.

The identification and localization of errors is a core task in peer review, yet the exponential growth of scientific output has made it increasingly difficult for human reviewers to reliably detect errors given the limited pool of experts. Recent advances in Large Language Models (LLMs) have sparked interest in their potential to support such evaluation tasks, from academic peer review to automated scientific assessment. However, despite the growing use of LLMs in review systems, their capabilities to pinpoint errors remain underexplored. In this work, we introduce Fault Localization Across Writing in Science (FLAWS), an automated benchmark consisting of 713 paper-error pairs designed to evaluate how effectively LLMs detect errors that undermine key claims in research papers. We construct the benchmark by systematically inserting claim-invalidating errors into peer-reviewed papers using LLMs, paired with an automated evaluation metric that measures whether models can identify and localize these errors. Developing such a benchmark presents unique challenges that we overcome: ensuring that the inserted errors are well-defined, challenging, and relevant to the content of the paper, avoiding artifacts that would make identification trivial, and designing a scalable, automated evaluation metric. On the resulting benchmark, we evaluate five frontier LLMs: Claude Sonnet 4.5, DeepSeek Reasoner v3.1, Gemini 2.5 Pro, GPT 5, and Grok 4. Among these, GPT 5 is the top-performing model, achieving 39.1