Hema Yoganarasimhan

Zikun Ye, Hema Yoganarasimhan

Large Language Models can generate synthetic survey responses at low cost, but their accuracy varies unpredictably across questions. We study the design problem of allocating a fixed budget of human respondents across estimation tasks when cheap LLM predictions are available for every task. Our framework combines three components. First, building on Prediction-Powered Inference, we characterize a question-specific rectification difficulty that governs how quickly the estimator's variance decreases with human sample size. Second, we derive a closed-form optimal allocation rule that directs more human labels to tasks where the LLM is least reliable. Third, since rectification difficulty depends on unobserved human responses for new surveys, we propose a meta-learning approach, trained on historical data, that predicts it for entirely new tasks without pilot data. The framework extends to general M-estimation, covering regression coefficients and multinomial logit partworths for conjoint analysis. We validate the framework on two datasets spanning different domains, question types, and LLMs, showing that our approach captures 61-79% of the theoretically attainable efficiency gains, achieving 11.4% and 10.5% MSE reductions without requiring any pilot human data for the target survey.

Amandeep Singh, Ye Liu, Hema Yoganarasimhan

Choice modeling is at the core of understanding how changes to the competitive landscape affect consumer choices and reshape market equilibria. In this paper, we propose a fundamental characterization of choice functions that encompasses a wide variety of extant choice models. We demonstrate how non-parametric estimators like neural nets can easily approximate such functionals and overcome the curse of dimensionality that is inherent in the non-parametric estimation of choice functions. We demonstrate through extensive simulations that our proposed functionals can flexibly capture underlying consumer behavior in a completely data-driven fashion and outperform traditional parametric models. As demand settings often exhibit endogenous features, we extend our framework to incorporate estimation under endogenous features. Further, we also describe a formal inference procedure to construct valid confidence intervals on objects of interest like price elasticity. Finally, to assess the practical applicability of our estimator, we utilize a real-world dataset from S. Berry, Levinsohn, and Pakes (1995). Our empirical analysis confirms that the estimator generates realistic and comparable own- and cross-price elasticities that are consistent with the observations reported in the existing literature.

Mehrzad Khosravi, Hema Yoganarasimhan

Search engines increasingly display LLM-generated answers shown above organic links, shifting search from link lists to answer-first summaries. Publishers contend these summaries substitute for source pages and cannibalize traffic, while platforms argue they are complementary by directing users through included links. We estimate the causal impact of Google's AI Overview (AIO) on Wikipedia traffic by leveraging the feature's staggered geographic rollout and Wikipedia's multilingual structure. Using a difference-in-differences design, we compare English Wikipedia articles exposed to AIO to the same underlying articles in language editions (Hindi, Indonesian, Japanese, and Portuguese) that were not exposed to AIO during the observation period. Across 161,382 matched article-language pairs, AIO exposure reduces daily traffic to English articles by approximately 15%. Effects are heterogeneous: relative declines are largest for Culture articles and substantially smaller for STEM, consistent with stronger substitution when short synthesized answers satisfy informational intent. These findings provide early causal evidence that generative-answer features in search engines can materially reallocate attention away from informational publishers, with implications for content monetization, search platform design, and policy.

Enoch Hyunwook Kang, Hema Yoganarasimhan

Large Language Models (LLMs) have recently enabled self-improving AI, i.e., AI that iteratively generates, evaluates, and refines its own outcomes. Recent studies have shown that self-improving AI focusing on prompt optimization can outperform state-of-the-art reinforcement-learning fine-tuned LLMs. Here, their `performance' is typically measured by query efficiency - the number of LLM-generated solution samples required to meet a certain performance threshold. However, in many societal applications, the primary limitation is not generating new solutions but evaluating them. For instance, evaluating an ad's effectiveness requires significant human feedback, which is far more costly and time-consuming than generating a candidate ad. To optimize for the evaluation efficiency objective, a natural approach is to extend Bayesian Optimization (BO), a framework proven optimal for evaluation efficiency, to the language domain. However, the difficulty of directly estimating suitable acquisition functions in LLMs' minds makes this extension challenging. This paper overcomes this challenge by proving that the combination of the simple and widely used Best-of-N selection strategy and simple textual gradients (i.e., textual edits from a critic model) statistically emulates the behavior of the gradients on the canonical UCB acquisition function, which induces optimal exploration in terms of evaluation efficiency. Based on this result, we propose TextGrad-Best-of-N Bayesian Optimization (T-BoN BO), a simple and eval-efficient language-space Bayesian optimization framework for AI self-improvement. We also empirically validate T-BoN BO by applying it to automated ad alignment tasks for persona distribution, demonstrating its superior performance compared to popular state-of-the-art baselines.

Mehrzad Khosravi, Max Kleiman-Weiner, Hema Yoganarasimhan

Many consumer decisions are repeated choices under uncertainty. Standard models capture these decisions using Bayesian learning and dynamic programming: consumers update beliefs from feedback and use those beliefs to guide future choices. In many markets, however, learning does not restart when consumers enter a new context: prior experience with a brand, product, or provider can shape beliefs in later, related decisions. We study this cross-context knowledge transfer, or meta-learning, in sequential choice. We design a hierarchical laboratory task in which participants repeatedly choose among airlines across routes and observe noisy binary outcomes. Reduced-form evidence shows that participants improve not only within routes, but also across routes: they choose better airlines earlier in later routes and reduce pseudo-regret. To identify the mechanism behind this transfer, we compare human choices to a no-transfer benchmark and a fully integrated Bayesian meta-learning benchmark. In particular, we introduce a class of boundedly rational meta dynamic programming policies, BRMDP(D), that approximate full integration using a limited number of hyper-posterior draws, denoted by D. Trial-by-trial likelihood comparisons show that low-D boundedly rational meta-learning, especially BRMDP(1), fits participant behavior better than both no transfer and fully integrated Bayesian transfer. Consumers, therefore, transfer brand-level regularities across contexts, but through coarse representations of prior uncertainty. The findings imply that models of consumer learning should allow for approximate cross-context transfer, and that managerial counterfactuals based on either no-transfer or fully integrated learning can be misleading.

Zikun Ye, Hema Yoganarasimhan, Yufeng Zheng

Modern media firms require automated and efficient methods to identify content that is most engaging and appealing to users. Leveraging a large-scale dataset from Upworthy (a news publisher), which includes 17,681 headline A/B tests, we first investigate the ability of three pure-LLM approaches to identify the catchiest headline: prompt-based methods, embedding-based methods, and fine-tuned open-source LLMs. Prompt-based approaches perform poorly, while both OpenAI-embedding-based models and the fine-tuned Llama-3-8B achieve marginally higher accuracy than random predictions. In sum, none of the pure-LLM-based methods can predict the best-performing headline with high accuracy. We then introduce the LLM-Assisted Online Learning Algorithm (LOLA), a novel framework that integrates Large Language Models (LLMs) with adaptive experimentation to optimize content delivery. LOLA combines the best pure-LLM approach with the Upper Confidence Bound algorithm to allocate traffic and maximize clicks adaptively. Our numerical experiments on Upworthy data show that LOLA outperforms the standard A/B test method (the current status quo at Upworthy), pure bandit algorithms, and pure-LLM approaches, particularly in scenarios with limited experimental traffic. Our approach is scalable and applicable to content experiments across various settings where firms seek to optimize user engagement, including digital advertising and social media recommendations.

Enoch H. Kang, Hema Yoganarasimhan, Lalit Jain

We study the problem of estimating Dynamic Discrete Choice (DDC) models, also known as offline Maximum Entropy-Regularized Inverse Reinforcement Learning (offline MaxEnt-IRL) in machine learning. The objective is to recover reward or $Q^*$ functions that govern agent behavior from offline behavior data. In this paper, we propose a globally convergent gradient-based method for solving these problems without the restrictive assumption of linearly parameterized rewards. The novelty of our approach lies in introducing the Empirical Risk Minimization (ERM) based IRL/DDC framework, which circumvents the need for explicit state transition probability estimation in the Bellman equation. Furthermore, our method is compatible with non-parametric estimation techniques such as neural networks. Therefore, the proposed method has the potential to be scaled to high-dimensional, infinite state spaces. A key theoretical insight underlying our approach is that the Bellman residual satisfies the Polyak-Lojasiewicz (PL) condition -- a property that, while weaker than strong convexity, is sufficient to ensure fast global convergence guarantees. Through a series of synthetic experiments, we demonstrate that our approach consistently outperforms benchmark methods and state-of-the-art alternatives.

Zikun Ye, Hema Yoganarasimhan

Large Language Models (LLMs) are increasingly used in systems that retrieve and summarize content from multiple sources, such as search engines and AI assistants. While these systems enhance user experience through coherent summaries, they obscure the individual contributions of original content creators, raising concerns about credit attribution and compensation. We address the challenge of valuing individual documents used in LLM-generated summaries by proposing a Shapley value-based framework for fair document valuation. Although theoretically appealing, exact Shapley value computation is prohibitively expensive at scale. To improve efficiency, we develop Cluster Shapley, a simple approximation algorithm that leverages semantic similarity among documents to reduce computation while maintaining attribution accuracy. Using Amazon product review data, we empirically show that off-the-shelf Shapley approximations, such as Monte Carlo sampling and Kernel SHAP, perform suboptimally in LLM settings, whereas Cluster Shapley substantially improves the efficiency-accuracy frontier. Moreover, simple attribution rules (e.g., equal or relevance-based allocation), though computationally cheap, lead to highly unfair outcomes. Together, our findings highlight the potential of structure-aware Shapley approximations tailored to LLM summarization and offer guidance for platforms seeking scalable and fair content attribution mechanisms.

Mohammad Mosaffa, Omid Rafieian, Hema Yoganarasimhan

Political polarization is a significant issue in American politics, influencing public discourse, policy, and consumer behavior. While studies on polarization in news media have extensively focused on verbal content, non-verbal elements, particularly visual content, have received less attention due to the complexity and high dimensionality of image data. Traditional descriptive approaches often rely on feature extraction from images, leading to biased polarization estimates due to information loss. In this paper, we introduce the Polarization Measurement using Counterfactual Image Generation (PMCIG) method, which combines economic theory with generative models and multi-modal deep learning to fully utilize the richness of image data and provide a theoretically grounded measure of polarization in visual content. Applying this framework to a decade-long dataset featuring 30 prominent politicians across 20 major news outlets, we identify significant polarization in visual content, with notable variations across outlets and politicians. At the news outlet level, we observe significant heterogeneity in visual slant. Outlets such as Daily Mail, Fox News, and Newsmax tend to favor Republican politicians in their visual content, while The Washington Post, USA Today, and The New York Times exhibit a slant in favor of Democratic politicians. At the politician level, our results reveal substantial variation in polarized coverage, with Donald Trump and Barack Obama among the most polarizing figures, while Joe Manchin and Susan Collins are among the least. Finally, we conduct a series of validation tests demonstrating the consistency of our proposed measures with external measures of media slant that rely on non-image-based sources.

Hema Yoganarasimhan, Ebrahim Barzegary, Abhishek Pani

Free trial promotions, where users are given a limited time to try the product for free, are a commonly used customer acquisition strategy in the Software as a Service (SaaS) industry. We examine how trial length affect users' responsiveness, and seek to quantify the gains from personalizing the length of the free trial promotions. Our data come from a large-scale field experiment conducted by a leading SaaS firm, where new users were randomly assigned to 7, 14, or 30 days of free trial. First, we show that the 7-day trial to all consumers is the best uniform policy, with a 5.59% increase in subscriptions. Next, we develop a three-pronged framework for personalized policy design and evaluation. Using our framework, we develop seven personalized targeting policies based on linear regression, lasso, CART, random forest, XGBoost, causal tree, and causal forest, and evaluate their performances using the Inverse Propensity Score (IPS) estimator. We find that the personalized policy based on lasso performs the best, followed by the one based on XGBoost. In contrast, policies based on causal tree and causal forest perform poorly. We then link a method's effectiveness in designing policy with its ability to personalize the treatment sufficiently without over-fitting (i.e., capture spurious heterogeneity). Next, we segment consumers based on their optimal trial length and derive some substantive insights on the drivers of user behavior in this context. Finally, we show that policies designed to maximize short-run conversions also perform well on long-run outcomes such as consumer loyalty and profitability.