As AI systems become more capable, deceptive behaviors can undermine evaluation and mislead users at deployment. Recent work has shown that lie detectors can accurately classify deceptive behavior, but they are not typically used in the training pipeline due to concerns around contamination and objective hacking. We examine these concerns by incorporating a lie detector into the labelling step of LLM post-training and evaluating whether the learned policy is genuinely more honest, or instead learns to fool the lie detector while remaining deceptive. Using DolusChat, a novel 65k-example dataset with paired truthful/deceptive responses, we identify three key factors that determine the honesty of learned policies: amount of exploration during preference learning, lie detector accuracy, and KL regularization strength. We find that preference learning with lie detectors and GRPO can lead to policies which evade lie detectors, with deception rates of over 85%. However, if the lie detector true positive rate (TPR) or KL regularization is sufficiently high, GRPO learns honest policies. In contrast, off-policy algorithms (DPO) consistently lead to deception rates under 25% for realistic TPRs. Our results illustrate a more complex picture than previously assumed: depending on the context, lie-detector-enhanced training can be a powerful tool for scalable oversight, or a counterproductive method encouraging undetectable misalignment.

Selection bias poses a widely recognized challenge for unbiased evaluation and learning in many industrial scenarios. For example, in recommender systems, it arises from the users' selective interactions with items. Recently, doubly robust and its variants have been widely studied to achieve debiased learning of prediction models, however, all of them consider a simple exact matching scenario, i.e., the units (such as user-item pairs in a recommender system) are the same between the training and test sets. In practice, there may be limited or even no overlap in units between the training and test. In this paper, we consider a more practical scenario: the joint distribution of the feature and rating is the same in the training and test sets. Theoretical analysis shows that the previous DR estimator is biased even if the imputed errors and learned propensities are correct in this scenario. In addition, we propose a novel super-population doubly robust estimator (SuperDR), which can achieve a more accurate estimation and desirable generalization error bound compared to the existing DR estimators, and extend the joint learning algorithm for training the prediction and imputation models. We conduct extensive experiments on three real-world datasets, including a large-scale industrial dataset, to show the effectiveness of our method.

Fraud detection models in payment networks train on chargeback labels that are systematically biased. Every label must survive three sequential gates: authorization (declined transactions generate no labels), issuer reporting (unreported fraud is invisible), and delay (pending chargebacks are missing at training time). Labels that do arrive may be corrupted by first-party misuse or issuer misclassification. A companion paper [arXiv:2605.27557] proved that these four impairments impose a minimax lower bound on detection performance. This paper asks: can that bound be achieved? We formalize the observation pipeline as a sequential missing-data problem with three propensity stages and a corruption layer, and construct the Sequential Triply Robust (STR) estimator. The STR corrects for all four impairments simultaneously and achieves the semiparametric efficiency bound -- no estimator can have lower asymptotic variance. It is sequentially triply robust: at each gate, consistency requires only that either the propensity model or the outcome regression is correctly specified, not both. We provide corruption correction via noise-rate-adjusted pseudo-labels, empirical Bayes shrinkage to stabilize inverse-propensity weights for small issuers, a plug-in variance estimator yielding valid confidence intervals, and a Bernstein concentration inequality for finite-sample guarantees. On the operational side, we derive the optimal training delay -- the maturity window that minimizes the sum of label-quality loss and model staleness -- and prove that the STR permits training on data that is days old rather than months old, decoupling model freshness from the chargeback maturity cycle. The STR provably dominates naive chargeback-based training in mean squared error for any sample size.

Estimating how much an intervention helps a given individual the conditional average treatment effect (CATE) is increasingly central to decision-making in medicine, economics, and policy, where an estimate is most useful when accompanied by a calibrated uncertainty interval. We study the few-placebo regime, in which one treatment arm is much smaller than the other, as arises in unequal-allocation trials and small-holdout $A/B$ tests. The standard estimator in this setting is the X-Learner, and a natural way to obtain credible intervals is to make its second stage Bayesian. We show that these intervals under-cover: they contain the true effect less often than their nominal level. We trace this to a structural cause the X-Learner's regression target inherits the bias of a nuisance model fitted to the small arm, so the posterior is centered away from the true effect and we find that the standard remedy, regressing an orthogonal doubly-robust score, is also unreliable here, since the regime's limited overlap leaves the estimator either highly variable or, once stabilized, biased once more. Both consequences reflect a pattern that extends beyond causal inference: a separately estimated variance is attached to a point estimate of a hard-to-learn quantity, and the point estimate's bias is not captured by that variance. We propose GP-CATE, which models each arm's outcome surface with a Gaussian process, so the scarce arm's uncertainty enters the posterior directly rather than as an unmodelled bias. Across synthetic and semi-synthetic benchmarks, GP-CATE attains calibrated coverage where the estimators we compare against including Causal Forest and BART do not, at the cost of intervals that are appropriately wide when the data are uninformative.

When treatment effects are naturally expressed as ratios -- as in medicine, pricing, and marketing -- the ratio-based CATE $τ(x) = E[Y|W=1,X=x] / E[Y|W=0,X=x]$ is the appropriate estimand. Yet existing estimators either impose a log-linear parametric structure or apply generic regression without robustness guarantees for this functional. We introduce the Q-Learner, which decomposes $τ(x)$ into a product of two odds ratios, reducing ratio-CATE estimation for binary outcomes to two propensity classification tasks. We further derive doubly robust augmentations for both S/T- and Q-style ratio learners and characterize their distinct robustness properties. In benchmarks on seven RCT datasets, the Q-Learner is the most consistently competitive method in low-conversion regimes, where its propensity-only construction sidesteps the imbalanced regression that hurts outcome-based estimators. On four observational datasets, where propensity must be estimated and confounding cannot be ruled out, the DR learners introduced here decisively come out on top, making them practitioners' natural default for confounded observational data.

Understanding the effects of interventions is central to scientific progress, with randomized controlled trials (RCTs) regarded as the gold standard for causal inference in many applied fields. However, RCTs are costly, time-consuming, and often constrained by ethical or practical limitations, motivating the need for causal methods able to draw conclusions from observational data. While such data is collected at ever larger scale, making its use for causal inference is often hindered by the fact that not all variables affecting treatment allocation and the outcome are observed - an issue known as unobserved confounding. In this paper, we introduce a new study design called confounder detection via treatment intent. The idea is to query a human expert who makes treatment decisions, and ask them to compare pairs of units proposed by a principled matching strategy, with the goal of eliciting unobserved variables that explain why treatment decisions differ. We provide a theoretical basis for such a procedure, ascertaining conditions under which such a study design may elicit unobserved confounders. Building on this newly established foundations, we study treatment effects of interventions in the intensive care unit (ICU). First, we show empirical evidence strongly indicating that electronic health records (EHRs) collected in ICUs are subject to unobserved confounding. By using clinical text notes as a proxy for physicians' knowledge and leveraging natural language processing, we provide a proof of concept for our methodology in a semi-synthetic environment with a known ground truth.

Some existing estimators and regularizers attempt to achieve unbiased estimation to improve the predictive performance.

In recommender systems, the collected data used for training is always subject to selection bias, which poses a great challenge for unbiased learning.

Neural Information Processing Systems http://nips.cc/