From Models to Systems: A Comprehensive Fairness Framework for Compositional Recommender Systems

Fairness research in machine learning often centers on ensuring equitable performance of individual models. However, real-world recommendation systems are built on multiple models and even multiple stages, from candidate retrieval to scoring and serving, which raises challenges for responsible development and deployment. This system-level view, as highlighted by regulations like the EU AI Act, necessitates moving beyond auditing individual models as independent entities. We propose a holistic framework for modeling system-level fairness, focusing on the end-utility delivered to diverse user groups, and consider interactions between components such as retrieval and scoring models. We provide formal insights on the limitations of focusing solely on model-level fairness and highlight the need for alternative tools that account for heterogeneity in user preferences. To mitigate system-level disparities, we adapt closed-box optimization tools (e.g., BayesOpt) to jointly optimize utility and equity. We empirically demonstrate the effectiveness of our proposed framework on synthetic and real datasets, underscoring the need for a system-level framework.

LLM Embeddings Improve Test-time Adaptation to Tabular $Y|X$-Shifts

For tabular datasets, the change in the relationship between the label and covariates ($Y|X$-shifts) is common due to missing variables (a.k.a. confounders). Since it is impossible to generalize to a completely new and unknown domain, we study models that are easy to adapt to the target domain even with few labeled examples. We focus on building more informative representations of tabular data that can mitigate $Y|X$-shifts, and propose to leverage the prior world knowledge in LLMs by serializing (write down) the tabular data to encode it. We find LLM embeddings alone provide inconsistent improvements in robustness, but models trained on them can be well adapted/finetuned to the target domain even using 32 labeled observations. Our finding is based on a comprehensive and systematic study consisting of 7650 source-target pairs and benchmark against 261,000 model configurations trained by 22 algorithms. Our observation holds when ablating the size of accessible target data and different adaptation strategies. The code is available at https://github.com/namkoong-lab/LLM-Tabular-Shifts.

QGym: Scalable Simulation and Benchmarking of Queuing Network Controllers

Queuing network control determines the allocation of scarce resources to manage congestion, a fundamental problem in manufacturing, communications, and healthcare. Compared to standard RL problems, queueing problems are distinguished by unique challenges: i) a system operating in continuous time, ii) high stochasticity, and iii) long horizons over which the system can become unstable (exploding delays). To spur methodological progress tackling these challenges, we present an open-sourced queueing simulation framework, QGym, that benchmark queueing policies across realistic problem instances. Our modular framework allows the researchers to build on our initial instances, which provide a wide range of environments including parallel servers, criss-cross, tandem, and re-entrant networks, as well as a realistically calibrated hospital queuing system. QGym makes it easy to compare multiple policies, including both model-free RL methods and classical queuing policies. Our testbed complements the traditional focus on evaluating algorithms based on mathematical guarantees in idealized settings, and significantly expands the scope of empirical benchmarking in prior work. QGym code is open-sourced at https://github.com/namkoong-lab/QGym.

PersonalLLM: Tailoring LLMs to Individual Preferences

As LLMs become capable of complex tasks, there is growing potential for personalized interactions tailored to the subtle and idiosyncratic preferences of the user. We present a public benchmark, PersonalLLM, focusing on adapting LLMs to provide maximal benefits for a particular user. Departing from existing alignment benchmarks that implicitly assume uniform preferences, we curate open-ended prompts paired with many high-quality answers over which users would be expected to display heterogeneous latent preferences. Instead of persona-prompting LLMs based on high-level attributes (e.g., user's race or response length), which yields homogeneous preferences relative to humans, we develop a method that can simulate a large user base with diverse preferences from a set of pre-trained reward models. Our dataset and generated personalities offer an innovative testbed for developing personalization algorithms that grapple with continual data sparsity--few relevant feedback from the particular user--by leveraging historical data from other (similar) users. We explore basic in-context learning and meta-learning baselines to illustrate the utility of PersonalLLM and highlight the need for future methodological development. Our dataset is available at https://huggingface.co/datasets/namkoong-lab/PersonalLLM

Differentiable Discrete Event Simulation for Queuing Network Control

Queuing network control is essential for managing congestion in job-processing systems such as service systems, communication networks, and manufacturing processes. Despite growing interest in applying reinforcement learning (RL) techniques, queueing network control poses distinct challenges, including high stochasticity, large state and action spaces, and lack of stability. To tackle these challenges, we propose a scalable framework for policy optimization based on differentiable discrete event simulation. Our main insight is that by implementing a well-designed smoothing technique for discrete event dynamics, we can compute pathwise policy gradients for large-scale queueing networks using auto-differentiation software (e.g., Tensorflow, PyTorch) and GPU parallelization. Through extensive empirical experiments, we observe that our policy gradient estimators are several orders of magnitude more accurate than typical REINFORCE-based estimators. In addition, We propose a new policy architecture, which drastically improves stability while maintaining the flexibility of neural-network policies. In a wide variety of scheduling and admission control tasks, we demonstrate that training control policies with pathwise gradients leads to a 50-1000x improvement in sample efficiency over state-of-the-art RL methods. Unlike prior tailored approaches to queueing, our methods can flexibly handle realistic scenarios, including systems operating in non-stationary environments and those with non-exponential interarrival/service times.

AExGym: Benchmarks and Environments for Adaptive Experimentation

Innovations across science and industry are evaluated using randomized trials (a.k.a. A/B tests). While simple and robust, such static designs are inefficient or infeasible for testing many hypotheses. Adaptive designs can greatly improve statistical power in theory, but they have seen limited adoption due to their fragility in practice. We present a benchmark for adaptive experimentation based on real-world datasets, highlighting prominent practical challenges to operationalizing adaptivity: non-stationarity, batched/delayed feedback, multiple outcomes and objectives, and external validity. Our benchmark aims to spur methodological development that puts practical performance (e.g., robustness) as a central concern, rather than mathematical guarantees on contrived instances. We release an open source library, AExGym, which is designed with modularity and extensibility in mind to allow experimentation practitioners to develop custom environments and algorithms.

Mathematical Programming For Adaptive Experiments

Adaptive experimentation can significantly improve statistical power, but standard algorithms overlook important practical issues including batched and delayed feedback, personalization, non-stationarity, multiple objectives, and constraints. To address these issues, the current algorithm design paradigm crafts tailored methods for each problem instance. Since it is infeasible to devise novel algorithms for every real-world instance, practitioners often have to resort to suboptimal approximations that do not address all of their challenges. Moving away from developing bespoke algorithms for each setting, we present a mathematical programming view of adaptive experimentation that can flexibly incorporate a wide range of objectives, constraints, and statistical procedures. By formulating a dynamic program in the batched limit, our modeling framework enables the use of scalable optimization methods (e.g., SGD and auto-differentiation) to solve for treatment allocations. We evaluate our framework on benchmarks modeled after practical challenges such as non-stationarity, personalization, multi-objectives, and constraints. Unlike bespoke algorithms such as modified variants of Thomson sampling, our mathematical programming approach provides remarkably robust performance across instances.

Pre-training and in-context learning IS Bayesian inference a la De Finetti

Accurately gauging uncertainty on the underlying environment is a longstanding goal of intelligent systems. We characterize which latent concepts pre-trained sequence models are naturally able to reason with. We go back to De Finetti's predictive view of Bayesian reasoning: instead of modeling latent parameters through priors and likelihoods like topic models do, De Finetti has long advocated for modeling exchangeable (permutation invariant) sequences of observables. According to this view, pre-training autoregressive models formulates informed beliefs based on prior observations ("empirical Bayes"), and forward generation is a simulated instantiation of an environment ("posterior inference"). This connection allows extending in-context learning (ICL) beyond predictive settings, highlighting sequence models' ability to perform explicit statistical inference. In particular, we show the sequence prediction loss over exchangeable documents controls performance on downstream tasks where uncertainty quantification is key. Empirically, we propose and demonstrate several approaches for encoding exchangeability in sequence model architectures: data augmentation, regularization, and causal masking.

Evaluating Model Performance Under Worst-case Subpopulations

The performance of ML models degrades when the training population is different from that seen under operation. Towards assessing distributional robustness, we study the worst-case performance of a model over all subpopulations of a given size, defined with respect to core attributes Z. This notion of robustness can consider arbitrary (continuous) attributes Z, and automatically accounts for complex intersectionality in disadvantaged groups. We develop a scalable yet principled two-stage estimation procedure that can evaluate the robustness of state-of-the-art models. We prove that our procedure enjoys several finite-sample convergence guarantees, including dimension-free convergence. Instead of overly conservative notions based on Rademacher complexities, our evaluation error depends on the dimension of Z only through the out-of-sample error in estimating the performance conditional on Z. On real datasets, we demonstrate that our method certifies the robustness of a model and prevents deployment of unreliable models.

Design and Scheduling of an AI-based Queueing System

To leverage prediction models to make optimal scheduling decisions in service systems, we must understand how predictive errors impact congestion due to externalities on the delay of other jobs. Motivated by applications where prediction models interact with human servers (e.g., content moderation), we consider a large queueing system comprising of many single server queues where the class of a job is estimated using a prediction model. By characterizing the impact of mispredictions on congestion cost in heavy traffic, we design an index-based policy that incorporates the predicted class information in a near-optimal manner. Our theoretical results guide the design of predictive models by providing a simple model selection procedure with downstream queueing performance as a central concern, and offer novel insights on how to design queueing systems with AI-based triage. We illustrate our framework on a content moderation task based on real online comments, where we construct toxicity classifiers by finetuning large language models.