On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,695 papers · 148 categories

Trend · papers per month

count % of corpus

48 results for “Learning to Defer”

A Causal Framework for Evaluating Deferring Systemscs.LG

This paper evaluates deferring systems using causal inference.

problem Evaluating the impact of deferring systems on model accuracy is challenging.

method The paper uses a causal inference framework to evaluate deferring systems, distinguishing between scenarios with and without access to human predictions.

result The approach allows identifying causal effects of deferring strategies on predictive accuracy.

causal inference deferring systems machine learning human-machine collaboration regression discontinuity

2024-05-29abs ↗pdf ↗

Learning-to-defer for sequential medical decision-making under uncertaintycs.LG

Framework for deferring decisions to experts in sequential medical settings.

problem Myopic and non-adaptive decision-making by ML models in sequential medical contexts.

method Sequential Learning-to-Defer (SLTD) framework using model-based reinforcement learning.

result Adaptive deferral policy improves trade-off between long-term outcomes and deferral frequency.

Sequential decision-making Medical decision-making Learning-to-defer Reinforcement learning Uncertainty quantification

2021-09-13abs ↗pdf ↗

Learning to Partially Defer for Sequencesstat.ME

New L2D framework allows deferring specific parts of a sequence prediction to experts.

problem Current L2D methods defer entire predictions, which is not ideal for long sequences.

method Proposes token-level and one-time rejectors to defer specific outputs of a model prediction to experts.

result Granular deferrals achieve better cost-accuracy tradeoffs than whole deferrals.

Learning to Defer Sequence Prediction Expert Integration Token-Level Deferral One-Time Deferral

2025-02-03abs ↗pdf ↗

Principled Approaches for Learning to Defer with Multiple Expertscs.LG

Study on learning to defer with multiple experts using new surrogate losses.

problem Learning to defer with multiple experts in a machine learning context.

method Introducing a new family of surrogate losses for the multiple-expert setting, proving

H

-consistency bounds, and designing learning algorithms.

result Explicit guarantees for new learning to defer algorithms based on minimization of these surrogate losses.

learning to defer multiple experts surrogate losses H-consistency SVHN

2023-10-23abs ↗pdf ↗

Learning-to-Defer with Expert-Conditional Advicestat.ML

A new method for learning to defer decisions with expert advice improves over standard methods.

problem Learning to defer decisions with expert advice in systems where expert information can be modified after selection.

method An augmented surrogate that operates on the composite expert-advice action space, providing consistency guarantees and excess-risk bounds.

result The method improves over standard Learning-to-Defer and adapts its advice acquisition behavior to the cost regime.

Learning-to-Defer Expert Advice Consistency Guarantees Excess-Risk Bounds Adaptive Advice Acquisition

2026-03-15abs ↗pdf ↗

Why Ask One When You Can Ask

k

? Learning-to-Defer to the Top-

k

Expertscs.LG

Unified framework for deferring queries to top-k experts, improving accuracy-cost trade-offs.

problem Limitation of existing L2D frameworks to single-expert deferral.

method Top-

k

Learning-to-Defer framework, including adaptive Top-

k(x)

variant.

result Superior accuracy-cost trade-offs with multi-expert deferral.

Learning-to-Defer Top-k Experts Multi-Expert Collaboration Surrogate Loss Bayes-consistency

2025-04-17abs ↗pdf ↗

Predict Responsibly: Improving Fairness and Accuracy by Learning to Deferstat.ML

In many machine learning applications, there are multiple decision-makers involved, both automated and human. The interaction between these agents often goes unaddressed in algorithmic development. In this work, we explore a simple version of this interaction with a two-stage framework containing an automated model and…

2017-11-17abs ↗pdf ↗

Realizable

H

-Consistent and Bayes-Consistent Loss Functions for Learning to Defercs.LG

Study of loss functions for learning to defer, proving consistency.

problem Learning to defer in machine learning.

method Introduced a family of surrogate losses parameterized by

Ψ

and proved their consistency.

result Proved realizable

H

-consistency and Bayes-consistency of specific surrogate losses.

learning to defer surrogate losses H-consistency Bayes-consistency non-increasing function $Ψ$

2024-07-18abs ↗pdf ↗

Optimized Deferral for Imbalanced Settingscs.LG

Optimized deferral improves accuracy in imbalanced settings.

problem Imbalance in expert predictions leads to suboptimal performance in two-stage learning to defer.

method Developed novel cost-sensitive learning algorithms and margin-based loss functions tailored for expert imbalance.

result MILD algorithm shows clear improvements over baselines in image classification and LLM routing tasks.

learning to defer cost-sensitive learning expert imbalance margin-based loss two-stage learning

2026-04-30abs ↗pdf ↗

Learning to Defer to Multiple Experts: Consistent Surrogate Losses, Confidence Calibration, and Conformal Ensemblesstat.ML

Study on learning to defer to multiple experts with consistent surrogates and confidence calibration.

problem Addressing the open problems of consistent surrogates, confidence calibration, and ensembling of experts.

method Derive two consistent surrogates (softmax and OvA) and propose a conformal inference technique for choosing experts.

result The OvA-based loss does not cause mis-calibration propagation, while the softmax-based loss does.

learning to defer consistent surrogates confidence calibration ensemble methods galaxy classification

2022-10-30abs ↗pdf ↗

Online Learning-to-Defer with Varying Expertsstat.ML

Online L2D algorithm for multiclass classification with varying experts.

problem Handling streaming data, changing expert availability, and shifting expert distribution.

method First online L2D algorithm with

O((n+n_e)T^{2/3})

and

O((n+n_e)\sqrt{T})

regret guarantees.

result Effective extension of standard L2D to settings with varying expert availability and reliability.

Learning-to-Defer Online Learning Varying Experts Bandit Feedback Multiclass Classification

2026-05-12abs ↗pdf ↗

Consistent Estimators for Learning to Defer to an Expertcs.LG

Paper proposes consistent estimators for learning to defer decisions to experts.

problem Learning algorithms often ignore expert decision-making in practical scenarios.

method Reduction to cost sensitive learning, novel surrogate loss for consistent estimation.

result Effective approach demonstrated on various tasks, showing consistency.

learning algorithms expert decision-making cost sensitive learning consistent estimation

2020-06-02abs ↗pdf ↗

Density-Ratio Losses for Post-Hoc Learning to Deferstat.ML

Study post-hoc Learning to Defer using density-ratio losses.

problem Optimizing decision-making between models and experts.

method Density-ratio losses for post-hoc L2D scorers, derived from class-probability estimation.

result The approach recovers known results and introduces new connections to expert comparison and anomaly detection.

Learning to Defer Density-ratio Post-hoc Chow's rule Bayes posterior

2026-05-19abs ↗pdf ↗

Adversarial Robustness in One-Stage Learning-to-Deferstat.ML

New framework improves adversarial robustness in one-stage L2D.

problem Adversarial robustness in one-stage Learning-to-Defer (L2D).

method Formalizes attacks, proposes cost-sensitive adversarial surrogate losses, establishes theoretical guarantees.

result Improves robustness against untargeted and targeted attacks while preserving clean performance.

Learning-to-Defer Adversarial Robustness One-Stage Cost-Sensitive Theoretical Guarantees

2025-10-13abs ↗pdf ↗

Calibrated Learning to Defer with One-vs-All Classifierscs.LG

A new L2D system produces calibrated probabilities of expert correctness without sacrificing accuracy.

problem Calibration of learning to defer systems for safety.

method One-vs-all classifiers with a consistent surrogate loss function.

result Proposes a calibrated L2D system that outperforms existing methods in accuracy and calibration.

learning to defer calibration one-vs-all classifiers surrogate loss AI safety

2022-02-08abs ↗pdf ↗

Learning to Defer for Causal Discovery with Imperfect Expertscs.LG

L2D-CD learns to defer expert recommendations in causal discovery.

problem Combining expert knowledge with data-driven results in causal discovery when expert recommendations may contradict data.

method Adapting learning-to-defer algorithms for pairwise causal discovery, L2D-CD learns a deferral function to select between expert recommendations and data-driven methods.

result L2D-CD outperforms both causal discovery methods and the expert used in isolation, identifying domains where the expert's performance is strong or weak.

causal discovery expert recommendations learning to defer pairwise causal discovery superior performance

2025-02-18abs ↗pdf ↗

One-Stage Top-

k

Learning-to-Defer: Score-Based Surrogates with Theoretical Guaranteesstat.ML

Unified model for prediction and deferral selects top-k entities efficiently.

problem Efficiently selecting top-k entities for deferral in machine learning.

method One-stage Top-

k

Learning-to-Defer framework with a convex surrogate.

result Unified model achieves superior accuracy-cost trade-offs.

Learning-to-Defer Top-k Score-based Convex surrogate Bayes-consistent

2025-05-15abs ↗pdf ↗

Achieving a Given Financial Goal with Optimal Deferred Term Insurance Purchasing Policyq-fin.PM

The paper optimizes insurance purchases for financial goals.

problem Maximizing probability of achieving financial goals with insurance.

method Analyzes deferred term insurance in deterministic and stochastic frameworks, considering income, consumption, and risky investment.

result Provides optimal insurance and investment strategies for achieving financial goals.

deferred term insurance financial planning optimal strategy stochastic framework death risk

2022-12-09abs ↗pdf ↗

Optimal Deferred Life Annuities to Minimize the Probability of Lifetime Ruinmath.OC

We find the minimum probability of lifetime ruin of an investor who can invest in a market with a risky and a riskless asset and can purchase a deferred annuity. Although we let the admissible set of strategies of annuity purchasing process to be increasing adapted processes, we find that the individual will not buy a …

2007-03-29abs ↗pdf ↗

Learning to Defer to a Population: A Meta-Learning Approachcs.LG

Framework allows systems to defer difficult decisions to unknown experts.

problem Ensuring safety and robustness in autonomous systems with unknown human experts.

method Meta-learning approach that adapts to new experts at test-time using optimization and model-based attention mechanisms.

result Framework can quickly adapt deferral policies to new experts without re-training.

meta-learning delegation autonomous systems human experts model-based

2024-03-05abs ↗pdf ↗

Adversarial Robustness in Two-Stage Learning-to-Defer: Algorithms and Guaranteesstat.ML

SARD improves adversarial robustness in two-stage L2D systems.

problem Adversarial attacks can manipulate query allocation in two-stage L2D systems.

method Introduces SARD, a convex learning algorithm with provable guarantees.

result SARD significantly improves robustness under adversarial attacks while maintaining strong clean performance.

Adversarial Robustness Two-Stage Learning-to-Defer SARD Surrogate Losses Bayes-Consistency

2025-02-03abs ↗pdf ↗

Beyond Augmented-Action Surrogates for Multi-Expert Learning-to-Deferstat.ML

A new method for multi-expert learning-to-defer avoids optimization issues.

problem Optimization issues in multi-expert learning-to-defer systems.

method A decoupled surrogate with a softmax classifier head and independent sigmoid heads per expert.

result First multi-expert L2D guarantee with a constant not growing with the expert pool.

learning-to-defer multi-expert decoupled surrogate optimization issues CIFAR-10

2026-04-10abs ↗pdf ↗

When More Experts Hurt: Underfitting in Multi-Expert Learning to Defercs.LG

Multi-expert L2D underfits more severely, requiring new methods.

problem Underfitting in multi-expert L2D settings.

method PiCCE (Pick the Confident and Correct Expert), a surrogate-based method.

result PiCCE effectively reduces multi-expert L2D to a single-expert-like problem, resolving underfitting.

multi-expert learning underfitting Pick the Confident and Correct Expert surrogate-based method statistical consistency

2026-02-19abs ↗pdf ↗

Optimal Query Allocation in Extractive QA with LLMs: A Learning-to-Defer Framework with Theoretical Guaranteescs.CL

Optimal query allocation improves extractive QA efficiency with LLMs.

problem Efficiency and reliability in extractive question answering with LLMs.

method Learning-to-Defer framework that allocates queries to specialized models.

result Enhanced answer reliability with reduced computational overhead.

extractive QA LLMs query allocation efficiency reliability

2024-10-21abs ↗pdf ↗

Calibrating conditional riskcs.LG

Study on estimating conditional risk in machine learning.

problem Estimating expected loss of prediction models given input features.

method Analyzed in classification and regression settings, showing equivalence to standard regression. Developed theoretical insights and empirical validation.

result Conditional risk calibration is distinct from existing uncertainty quantification problems.

conditional risk machine learning uncertainty quantification classification regression

2026-04-22abs ↗pdf ↗

Multistage Defer Trees for Hybrid Interpretability: If at First You Can't Succeed, Tree Againcs.LG

Multistage Defer Trees improve model accuracy while maintaining interpretability.

problem Balancing model accuracy and interpretability, especially in noisy domains.

method A sequence of sparse decision trees that defer predictions to the next tree or a black box.

result Matches the performance of complex tree-based ensembles while using only one or a few sparse trees.

Decision Trees Interpretability Accuracy Ensemble Methods Sparse Models

2026-06-30abs ↗pdf ↗

Learning to Make Adherence-Aware Advicestat.ML

The paper tackles AI advice giving by considering adherence levels and defer options.

problem Inadequate consideration of human adherence to AI recommendations.

method Sequential decision-making model that considers adherence levels and incorporates a defer option.

result Specialized learning algorithms provide better convergence and empirical performance.

adherence-aware advice sequential decision-making reinforcement learning human-AI interactions

2023-10-01abs ↗pdf ↗

Regression with Multi-Expert Deferralcs.LG

A novel framework for regression with multiple experts, addressing challenges in infinite and continuous label spaces.

problem Challenges in regression with multiple experts due to the infinite and continuous nature of the label space.

method Introduces a novel framework for regression with deferral, analyzing both single-stage and two-stage scenarios with new surrogate loss functions.

result Proves

H

-consistency bounds for both single-stage and two-stage methods, providing stronger guarantees than Bayes consistency.

Regression Multi-Expert Deferral Consistency Bounds Surrogate Loss Functions Infinite Label Space

2024-03-28abs ↗pdf ↗

Optimal allocation to deferred income annuitiesq-fin.PM

The paper optimizes DIA purchase policies using lifecycle models and asset allocation.

problem Determining the optimal allocation to Deferred Income Annuities (DIAs).

method Employed a lifecycle model with utility of consumption and bequest, formalized optimization process, analyzed results, and extended model to include asset allocation.

result Optimal DIA allocation varies based on refundability, asset allocation, and perceived longevity.

Deferred Income Annuities Lifecycle Model Optimization Utility of Consumption Bequest

2021-11-01abs ↗pdf ↗

A Two-Stage Learning-to-Defer Approach for Multi-Task Learningstat.ML

A new framework integrates classification and regression tasks in multi-task learning.

problem Jointly solving classification and regression tasks in multi-task scenarios.

method Two-Stage Learning-to-Defer (L2D) framework with a unified deferral mechanism.

result Unified deferral mechanism ensures convergence to the Bayes-optimal rejector.

multi-task learning classification regression Bayes-consistency L2D

2024-10-21abs ↗pdf ↗

Donor's Deferral and Return Behavior: Partial Identification from a Regression Discontinuity Design with Manipulationstat.ME

Volunteer labor can temporarily yield lower benefits to charities than its costs. In such instances, organizations may wish to defer volunteer donations to a later date. Exploiting a discontinuity in blood donations' eligibility criteria, we show that deferring donors reduces their future volunteerism. In our setting, …

2019-10-04abs ↗pdf ↗

Learning What to Defer for Maximum Independent Setscs.LG

A new DRL scheme optimizes solving large graphs' maximum independent set problem.

problem Efficiently solving maximum independent set problems on large graphs.

method Learning what to defer (LwD) to adaptively control the number of stages.

result Significantly outperforms state-of-the-art DRL and conventional solvers.

reinforcement learning maximum independent set graph optimization deep learning adaptive algorithms

2020-06-17abs ↗pdf ↗

From Soft Classifiers to Hard Decisions: How fair can we be?cs.LG

A popular methodology for building binary decision-making classifiers in the presence of imperfect information is to first construct a non-binary "scoring" classifier that is calibrated over all protected groups, and then to post-process this score to obtain a binary decision. We study the feasibility of achieving vari…

2018-10-03abs ↗pdf ↗

Mastering Multiple-Expert Routing: Realizable

H

-Consistency and Strong Guarantees for Learning to Defercs.LG

This paper tackles deferral learning with multiple experts, providing strong theoretical guarantees.

problem Optimizing input assignment to experts balancing accuracy and computational cost.

method Introducing new surrogate loss functions and efficient algorithms with strong theoretical learning guarantees.

result Realizable

H

-consistency,

H

-consistency bounds, and Bayes-consistency for deferral learning.

deferral learning multiple experts surrogate loss functions theoretical guarantees natural language generation

2025-06-25abs ↗pdf ↗

When to Accept Automated Predictions and When to Defer to Human Judgment?cs.LG

The paper proposes a method to assess when automated predictions are reliable.

problem Ensuring reliability and safety of automated decision-making in machine learning.

method Clustering to measure distances between outputs and class centroids, defining a safety threshold based on these distances.

result The proposed metric can efficiently determine when automated predictions are acceptable and when they should be deferred.

automated decision-making reliability safety distribution shifts machine learning

2024-07-10abs ↗pdf ↗

Towards Unbiased and Accurate Deferral to Multiple Expertscs.LG

Improves accuracy and fairness in prediction systems with multiple domain experts.

problem Designing unbiased and accurate deferral systems with multiple experts.

method Proposes a framework for learning a classifier and deferral system that chooses to defer to multiple human experts.

result Significantly improves accuracy and fairness of final predictions compared to baselines.

deferral systems multiple experts machine learning accuracy fairness

2021-02-25abs ↗pdf ↗

Fast Catch-Up, Late Switching: Optimal Batch Size Scheduling via Functional Scaling Lawscs.LG

New framework optimizes deep learning training by deferring large batch sizes to late stages.

problem Optimizing batch size scheduling for deep learning training efficiency.

method Introduced the functional scaling law (FSL) framework to analyze and optimize batch size scheduling.

result Large batch sizes can be deferred to late training stages without sacrificing performance.

batch size scheduling deep learning functional scaling law large language models late switching

2026-02-15abs ↗pdf ↗

A synthetic Gannon-Lee incompleteness theoremgr-qc

Synthetic proof of Gannon-Lee theorem for spacetimes.

problem Proving incompleteness in globally hyperbolic spacetimes.

method Synthetic null energy condition and synthetically asymptotically regular trappedness condition.

result Generalized classical incompleteness theorem to weighted spacetimes.

Gannon-Lee theorem spacetimes incompleteness null energy condition trappedness

2026-02-15abs ↗pdf ↗

Bifurcation in optimal retirementq-fin.PM

Study shows bifurcation in optimal retirement planning.

problem Optimal consumption and retirement planning model.

method Cobb-Douglas utility, simple model with wealth bifurcation.

result Critical wealth level leads to a continuum of retirement trajectories.

retirement planning utility maximization Cobb-Douglas bifurcation consumption decisions

2025-06-02abs ↗pdf ↗

Information Forestscs.LG

We describe Information Forests, an approach to classification that generalizes Random Forests by replacing the splitting criterion of non-leaf nodes from a discriminative one -- based on the entropy of the label distribution -- to a generative one -- based on maximizing the information divergence between the class-con…

2012-02-07abs ↗pdf ↗

Diffusion Boosted Treesstat.ML

DBT combines diffusion models and boosting for supervised learning.

problem Supervised learning problems.

method Diffusion Boosting paradigm and Diffusion Boosted Trees (DBT).

result DBT outperforms deep neural network-based diffusion models and is effective on real-world classification tasks.

supervised learning diffusion models boosting decision trees conditional distributions

2024-06-03abs ↗pdf ↗

Learning Multimodal Transition Dynamics for Model-Based Reinforcement Learningstat.ML

In this paper we study how to learn stochastic, multimodal transition dynamics in reinforcement learning (RL) tasks. We focus on evaluating transition function estimation, while we defer planning over this model to future work. Stochasticity is a fundamental property of many task environments. However, discriminative f…

2017-05-01abs ↗pdf ↗

Householder Dice: A Matrix-Free Algorithm for Simulating Dynamics on Gaussian and Random Orthogonal Ensemblescs.IT

HD algorithm simulates dynamics on random matrix ensembles without generating full matrices.

problem Simulating dynamics on dense random matrix ensembles with high space and time complexity.

method Householder reflectors for adaptive and recursive construction, deferring decisions.

result Significant reductions in runtime and memory footprint for practical

T \ll n

Householder reflectors random matrix ensembles deferring decisions simulation high-dimensional systems

2021-01-19abs ↗pdf ↗

Variational Gaussian Approximation in Replica Analysis of Parametric Modelscond-mat.dis-nn

We use variational Gaussian approximations to analyze parametric models with unknown data-generating distributions.

problem Analyzing inference and learning in parametric models with unknown or intractable data-generating distributions.

method Replica method with variational Gaussian approximation in grand canonical formalism.

result Stationarity conditions adaptively determine parameters of the trial Hamiltonian for each dataset.

replica method variational Gaussian approximation parametric models grand canonical formalism learning curves

2025-09-15abs ↗pdf ↗

Quantum Cross Entropy and Maximum Likelihood Principlequant-ph

Quantum machine learning uses quantum cross entropy to minimize loss, but measurement loss affects this process.

problem Quantum machine learning's loss minimization through cross entropy is affected by measurement outcomes.

method Defined quantum cross entropy, proved its lower bounds, and investigated its relation to quantum fidelity and likelihood.

result Quantum cross entropy is lower-bounded by negative log-likelihood when derived from quantum data, but measurement outcomes can cause loss.

quantum machine learning cross entropy maximum likelihood quantum fidelity measurement loss

2021-02-23abs ↗pdf ↗

Learning Imbalanced Datasets with Label-Distribution-Aware Margin Losscs.LG

Deep learning algorithms can fare poorly when the training dataset suffers from heavy class-imbalance but the testing criterion requires good generalization on less frequent classes. We design two novel methods to improve performance in such scenarios. First, we propose a theoretically-principled label-distribution-awa…

2019-06-18abs ↗pdf ↗

Strategic Collusion of LLM Agents: Market Division in Multi-Commodity Competitionscs.GT

LLMs can collude in market divisions, maximizing profits.

problem Strategic collusion of LLM agents in multi-commodity markets.

method Examined LLMs in Cournot competition frameworks, analyzing pricing and resource allocation strategies.

result LLMs can monopolize specific commodities without direct human input or explicit collusion commands.

LLM collusion market division machine learning Cournot competition AI pricing

2024-09-19abs ↗pdf ↗

Ricci flow with surgery on three-manifoldsmath.DG

This is a technical paper, which is a continuation of math.DG/0211159. Here we construct Ricci flow with surgeries and verify most of the assertions, made in section 13 of that e-print; the exceptions are (1) the statement that manifolds that can collapse with local lower bound on sectional curvature are graph manifold…

2003-03-10abs ↗pdf ↗