Transforms any test into anytime-valid with sample savings.
problem Sequential data invalidates classical test guarantees.
method Predicts test outcomes to create anytime-valid stopping rules.
result Ensures Type-I error control and near-optimal power.
Paper offers anytime-valid inference for causal parameters using DML.
problem Classic DML is only valid asymptotically for a fixed sample size.
method Time-uniform DML results for anytime-valid inference.
result Valid inference at any arbitrary stopping time.
Adaptive auditing improves AI robustness testing with anytime-valid guarantees.
problem Cost and time of annotation limit rigorous AI failure mode characterization.
method Introduces hypothesis testing framework for adaptive audits using SAVI.
result Proves anytime-valid type-I error control and robustness certification.
Develops an anytime-valid framework for optimal policy identification from logged contextual bandit data.
problem Selecting the optimal policy from a candidate policy class while monitoring evidence continuously.
method Constructs a time-indexed set that retains the true optimal policy set uniformly over time.
result The procedure allows the analyst to monitor policy values, eliminate clearly suboptimal policies, and stop at data-dependent times without invalidating inference.
GAAVI offers anytime-valid tests for CMF global null and contrasts.
problem Inference on the conditional mean function for high confidence decisions.
method Asymptotic anytime-valid tests for CMF global null and contrasts.
result Achieves asymptotic type-I error guarantees, power one, and optimal sample complexity.
Develops anytime-valid stopping rules for SGD based on observed trajectory.
problem Stopping stochastic gradient descent (SGD) based on observed trajectory.
method Develops anytime-valid confidence sequences for stochastic gradient methods.
result Statistically valid, time-uniform stopping rules for SGD across convex and nonconvex settings.
Four geometries govern sequential and distribution-free inference.
problem Sequential and distribution-free inference challenges.
method Four distinct admissibility geometries.
result Four classes of admissible procedures are pairwise non-nested.
CITE algorithm provides anytime-valid certification of model outputs.
problem Challenges in controlling error levels in LLM self-consistency.
method Certification by Intersection-union Testing with E-processes (CITE) algorithm.
result Provable control of false certification at any prescribed level under arbitrary stopping rules.
Anytime-valid confirmation of label-shift corrections
problem Small-batch scientific deployments with scarce labeled outcomes
method Conditional e-value and martingale-based rule
result Nonnegative martingale and anytime-valid confirmation rule
Paper develops a new watermarking framework for LLMs.
problem Efficiently distinguishing machine-generated from human text.
method Develops Anchored E-Watermarking framework for anytime-valid inference.
result Framework enables valid, anytime-inference with reduced sample efficiency.
Improves decision tree performance by correcting split selection errors.
problem Invalid statistical guarantees in split selection for decision trees.
method Introduces anytime-valid inference to provide valid statistical guarantees.
result Provides anytime-valid control of false splits under arbitrary data streams.
Paper improves PAC-Bayes bounds for various loss types.
problem Improving PAC-Bayes bounds for different types of losses.
method Introducing new high-probability PAC-Bayes bounds for bounded and general tail behaviors losses, and extending to anytime-valid bounds.
result New fast-rate and mixed-rate bounds for losses with bounded ranges, and parameter-free bounds for losses with general tail behaviors.
PITMonitor monitors model calibration over time with formal error guarantees.
problem Fixed-sample tests applied to models over time can lead to false alarms.
method PITMonitor uses mixture e-processes to detect distributional shifts in probability integral transforms.
result PITMonitor achieves competitive detection rates on river's FriedmanDrift benchmark.
E-C2ST uses E-values for high-dimensional data two-sample tests.
problem Statistical testing for high-dimensional data.
method Combines split likelihood ratio tests and predictive independence tests, using E-values for anytime-valid sequential tests.
result E-C2ST achieves enhanced statistical power by partitioning datasets into multiple batches.
Extends risk control to adaptive data collection, anytime-valid guarantees.
problem Ensuring safety of machine learning models with critical risk measures.
method Sequential risk controlling prediction sets (RCPS) for adaptive data collection and active labeling.
result Anytime-valid guarantees for risk control in sequential data collection.
Develops anytime-valid conformal and PAC prediction for streaming data.
problem Lack of guarantee in traditional conformal methods for sequential settings.
method Extends conformal and PAC prediction frameworks to handle streaming data.
result Provides anytime-valid prediction sets for sequential settings.
New methods for anytime-valid off-policy inference in contextual bandits.
problem Estimating properties of hypothetical policies in adaptive experiments.
method Modern martingale techniques for comprehensive OPE inference.
result Valid anytime inference for off-policy mean reward values and entire reward distributions.
Extends FC-RAG to anytime-valid sequential coverage for language model swarms.
problem Maintain distribution-free coverage for a swarm of weak language models over time.
method Introduces Anytime-FC-RAG, a sequential extension with a summable calibration-deviation budget.
result Achieves time-uniform alarm validity and safety under predictable adaptive control.
This work creates a CS for non-negative heavy-tailed data with bounded mean.
problem Constructing a confidence sequence for non-negative heavy-tailed data with bounded mean.
method Non-parametric, non-asymptotic lower confidence sequence construction.
result The constructed CS is efficient and can be converted into a closed-interval CS.
Efficiently audits model fairness with continuous monitoring and flexible data collection.
problem Continuous monitoring and flexible data collection for fairness auditing.
method Sequential, anytime-valid inference and game-theoretic statistics.
result Demonstrated efficacy on three fairness datasets.
CSA fills a gap in RLVR-trained LLM deployment by providing anytime-valid selective risk control.
problem Deployment of RLVR-trained LLMs in regulated organizations requires a safety certificate for every round without waiting for long-run averages.
method CSA uses a (test statistic, validity guarantee, deployment rule) framework to fill the gap, maintaining a Ville-type e-process per threshold on a Bonferroni grid.
result CSA provides the first anytime-valid selective risk control for RLVR-trained LLMs, matching the long-run average certification rate and satisfying pathwise validity and non-refusing deployment on every cell.
Predictive e-values enhance statistical inference across various tasks.
problem Insufficient data limits traditional statistical inference.
method Apply prediction-powered inference to e-values.
result Every e-value-based inference has a prediction-powered counterpart.
A new method detects distribution shifts faster than existing CTMs.
problem Detecting distribution shifts in data streams with contamination issues.
method Uses a fixed reference dataset to compare each new sample, avoiding contamination.
result Detects distribution shifts faster and more reliably than standard CTMs.
Unified framework for anytime-valid PAC-Bayes bounds.
problem Deriving time-uniform PAC-Bayes bounds for stochastic processes.
method Combines four tools: nonnegative supermartingales, method of mixtures, Donsker-Varadhan formula, and Ville's inequality.
result Unified PAC-Bayes theorem for a wide class of discrete stochastic processes.
This work optimizes identifying good arms in nonparametric multi-armed bandits.
problem Efficiently identifying arms with high means in nonparametric settings.
method Combining reward-maximizing sampling with a nonparametric sequential test for anytime-valid labeling.
result Achieves minimax optimal stopping times for identifying arms above a threshold.
BRACE addresses noncompliance in bandits, offering methods for recommendation and treatment policies.
problem Noncompliance in bandit problems complicates learning objectives and treatment effects.
method BRACE formalizes objective-choice, identifies direct-control regimes, and proposes a phase-doubling algorithm for IV inversion.
result BRACE delivers valid policy values and structural uncertainty, even under weak identification and homogeneity failure.
New flexible confidence sequences for robust statistical inference.
problem Creating robust statistical inference methods that work under mild assumptions.
method Proposed a new class of asymptotic time-uniform confidence sequences.
result Sharp asymptotic time-uniform confidence sequences achieved under mild assumptions.
Sequential tests for nonparametric hypotheses using supermartingales.
problem Designing valid sequential tests for nonparametric null hypotheses.
method Using elicitable and identifiable functionals, nonnegative supermartingales, and Online Convex Optimization.
result Rigorous guarantees on asymptotic power for a wide range of alternative hypotheses.
Extends PPI to sequential setting, improving inference over time.
problem Sequential data growth with unlabelled data.
method Prediction-powered confidence sequence procedures using Ville's inequality and the method of mixtures.
result Asymptotically valid uniformly over time, accommodating prior knowledge.
A universal framework for constructing confidence sets using sequential likelihood mixing.
problem Constructing reliable confidence sets for realizable likelihood functions.
method Sequential likelihood mixing, integrating Bayesian inference and regret inequalities.
result Establishes fundamental connections and provable coverage guarantees for various inference techniques.
Unified framework for online LLM watermark detection using e-processes.
problem Detecting AI-generated text from human-written content in online settings.
method Unified framework based on e-processes for anytime-valid hypothesis testing on independence.
result Proposed methods achieve competitive performance in watermark detection.
We propose a general method for constructing hypothesis tests and confidence sets that have finite sample guarantees without regularity conditions. We refer to such procedures as "universal." The method is very simple and is based on a modified version of the usual likelihood ratio statistic, that we call "the split li…
The paper extends conformal risk control to be valid with high probability over a growing calibration dataset.
problem Valid risk control over a growing calibration dataset.
method Quantile-based arguments for anytime-valid control.
result Guarantees remain valid with high probability over a cumulatively growing calibration dataset.
E-values enhance conformal prediction methods.
problem Distribution-free uncertainty quantification.
method Reformulation of conformal prediction using e-values.
result E-values offer new theoretical and practical capabilities.
M-FISHER detects and adapts to streaming data shifts with statistical validity and stability.
problem Detecting and adapting to distributional shifts in streaming data.
method Constructs an exponential martingale from non-conformity scores and applies Ville's inequality for detection. Fisher-preconditioned updates for adaptation.
result Establishes M-FISHER as a principled approach for robust, anytime-valid detection and geometrically stable adaptation.
The paper proves generalization bounds and stopping rules for self-selected data in reciprocal learning.
problem Generalization of learning algorithms using self-selected data.
method Proves universal generalization bounds using covering numbers and Wasserstein ambiguity sets.
result Provides stopping rules for reciprocal learning algorithms to ensure out-of-sample performance.
Efficient inference method for adaptive experiments with tighter confidence sequences.
problem Efficient inference of Average Treatment Effect in a changing policy sequential experiment.
method Semiparametric efficient inference using Adaptive Augmented Inverse-Probability Weighted estimator and asymptotic confidence sequences.
result Derives tighter confidence sequences for adaptive experiments under data-dependent stopping times.
The paper proposes an efficient method for estimating ATEs using adaptive experiments.
problem Estimating average treatment effects (ATEs) with minimal sample size and high accuracy.
method The paper defines and uses the efficient treatment-assignment probability to sequentially assign treatments, estimating ATEs using an Adaptive Augmented Inverse Probability Weighting (A2IPW) estimator.
result The proposed experimental design and A2IPW estimator achieve the minimized semiparametric efficiency bound and provide anytime valid confidence intervals for early stopping.
New method improves treatment effect estimation in adaptive experiments with noncompliance.
problem Estimating average treatment effect in adaptive experiments with binary instrumental variable.
method AMRIV estimator that balances outcome noise and compliance variability.
result AMRIV achieves semiparametric efficiency bound and is robust to noncompliance.
MAXMINLCB optimizes unknown target functions with preference feedback using a Stackelberg game approach.
problem Optimizing unknown target functions with pairwise comparisons and human feedback.
method MAXMINLCB, a zero-sum Stackelberg game, balances exploration and exploitation.
result MAXMINLCB consistently outperforms existing algorithms with a rate-optimal regret guarantee.
A new stopping rule based on E-values helps efficiently use sampling in Bayesian Deep Ensembles.
problem How long should sampling continue in Bayesian Deep Ensembles to yield significant improvements?
method Formulated as a sequential anytime-valid hypothesis test, using E-values to decide when to stop sampling.
result Only a fraction of the full-chain budget is often required for significant improvements.
EENNs improve inference efficiency but need nested prediction sets for reliable uncertainty estimates.
problem Non-nested prediction sets from standard uncertainty quantification methods in EENNs.
method Introduced anytime-valid confidence sequences (AVCSs) tailored for EENNs.
result AVCSs generate nested prediction sets across EENN exits, addressing the issue of non-nested sets.
Framework for online hypothesis testing across various data types.
problem Testing various nonparametric hypotheses in data streams.
method Unified framework using operators on data distributions, leveraging ML models.
result Efficient, adaptive, and error-controlled sequential tests.
In sequential anytime-valid inference, any admissible procedure must be based on e-processes: generalizations of test martingales that quantify the accumulated evidence against a composite null hypothesis at any stopping time. This paper proposes a method for combining e-processes constructed in different filtrations b…
Paper compares two forecasters using novel online inference methods.
problem How to compare forecasters without distributional assumptions.
method Confidence sequences and game-theoretic statistical framework for sequential testing.
result Valid methods for comparing forecasters without distributional assumptions.
Framework for online resource allocation using social welfare functions.
problem Optimal allocation of resources over time steps in a population.
method Confidence sequence framework for SWF-based online learning and inference, valid for any monotonic, concave, and Lipschitz-continuous SWF.
result Achieves near-optimal regret of i l d e O ( n + n k T ) ilde{O}(n+\sqrt{nkT}) i l d e O ( n + nk T ) for SWF-agnostic algorithm SWF-UCB. Algorithm identifies best arm with biased proxy and selective ground truth audits.
problem Fixed-confidence best-arm identification with biased proxy and selective ground truth.
method Propensity-weighted estimator and adaptive auditing algorithm.
result Plug-in Neyman rule achieves near-oracle audit efficiency.
Bayes-assisted confidence sequences improve efficiency for bounded means.
problem Efficient uncertainty quantification for bounded IID means without parametric assumptions.
method Bayesian working predictive model selects adaptive martingale updates maximizing predictive log-growth.
result Asymptotically log-optimal performance with informative priors reducing width and sampling effort.