IDT learns human preferences from uncertain decisions, even when humans are suboptimal.
problem Learning human preferences from uncertain and suboptimal decisions.
method Inverse decision theory (IDT) framework, statistical analysis of IDT, characterizing sample complexity.
result Learning preferences is easier when decisions are more uncertain, even if humans are suboptimal.
Paper tackles RLHF with DCPPO method, proving near-optimal suboptimality.
problem Challenges in offline RLHF with limited human feedback and bounded rationality.
method DCPPO method involving three stages: MLE, reward function recovery, and pessimistic value iteration.
result DCPPO's suboptimality almost matches classical pessimistic offline RL in terms of distribution shift and dimension.
Inferring intent from observed behavior has been studied extensively within the frameworks of Bayesian inverse planning and inverse reinforcement learning. These methods infer a goal or reward function that best explains the actions of the observed agent, typically a human demonstrator. Another agent can use this infer…
Privacy-preserving reinforcement learning from human feedback using decoupled reward modeling.
problem Training large language models with sensitive user information while preserving privacy.
method Proposes a privacy-preserving framework that imposes differential privacy on reward learning only.
result Privacy contributes an additional additive term to the suboptimality gap, and the upper bound is rate-optimal up to logarithmic factors.
Human irrationality can improve AI design, study shows.
problem Improving AI by learning from human decision-making biases.
method Developed a novel POMDP model to simulate human decision-making in contextual choice tasks.
result Reinforcement learners can exploit human irrationalities to make better decisions.
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.
Paper identifies problematic baselines in Shapley value explanations and proposes a reweighting mechanism.
problem Identifying and addressing the suboptimality of baselines in Shapley value feature importance analysis.
method Analyzed suboptimality of baselines, identified problematic baseline, generalized uninformativeness, and designed a reweighting mechanism.
result Proposed uncertainty-based reweighting mechanism effectively accelerates computation and improves explanation quality.
Trust-aware MAB improves learning performance by accounting for human deviation.
problem Learning performance suffers when humans deviate from recommended policies due to lack of trust.
method Integrates a dynamic trust model into MAB framework, establishing minimax regret and proposing a two-stage trust-aware procedure.
result Proves near-optimal statistical guarantees for trust-aware MAB algorithms.
Reasoning about graphs evolving over time is a challenging concept in many domains, such as bioinformatics, physics, and social networks. We consider a common case in which edges can be short term interactions (e.g., messaging) or long term structural connections (e.g., friendship). In practice, long term edges are oft…
Interpole learns transparent decision-making policies from data.
problem Understanding human decision-making in opaque environments.
method Interpole combines belief-update and belief-action mapping estimation.
result Interpole provides interpretable models of decision-making behavior.
This thesis tackles learning reward functions from human comparative feedback.
problem Designing reward functions for complex tasks is challenging and humans often provide suboptimal demonstrations.
method Proposes learning reward functions from comparative feedback (pairwise comparisons, best-of-many choices, rankings, scaled comparisons) and active learning techniques.
result Demonstrates the effectiveness of learning reward functions from comparative feedback in various domains.
RAFT fine-tunes models using high-quality samples to align them with human preferences.
problem Aligning generative models with human ethics and preferences.
method RAFT selects high-quality samples, discards undesired behavior, and fine-tunes the model on filtered samples.
result RAFT improves model performance in reward learning and automated metrics.
The purpose of this article is to propose a new "theory," the Strategic Analysis of Financial Markets (SAFM) theory, that explains the operation of financial markets using the analytical perspective of an enlightened gambler. The gambler understands that all opportunities for superior performance arise from suboptimal …
Gradient-based algorithm improves model performance with triage.
problem Improving model accuracy with human expert involvement.
method Formal characterization of triage, optimal triage policy as deterministic threshold rule, gradient-based algorithm.
result Gradient-based algorithm finds triage policies and models with increasing performance.
Two entropy measures quantify suboptimal portfolio performance.
problem Measuring suboptimality in investment portfolios.
method Relative entropy (KL divergence) calculations.
result Suboptimal portfolios appear better than Kelly portfolios under certain measures.
Novel IRL method identifies suboptimal medical decisions in ICU data.
problem Identifying suboptimal medical decisions in clinical settings.
method Incorporates Inverse Reinforcement Learning with a pruning step to identify and remove suboptimal actions.
result Pruning step effectively identifies clinical priorities and values from suboptimal data.
Recent work in reinforcement learning demonstrated that learning solely through self-play is not only possible, but could also result in novel strategies that humans never would have thought of. However, optimization methods cast as a game between two players require careful tuning to prevent suboptimal results. Hence,…
Integer programming (IP) is a general optimization framework widely applicable to a variety of unstructured and structured problems arising in, e.g., scheduling, production planning, and graph optimization. As IP models many provably hard to solve problems, modern IP solvers rely on many heuristics. These heuristics ar…
This work explores a social learning problem with agents having nonidentical noise variances and mismatched beliefs. We consider an N-agent binary hypothesis test in which each agent sequentially makes a decision based not only on a private observation, but also on preceding agents' decisions. In addition, the agents…
SharedRep-RLHF learns shared traits for diverse groups, improving fairness and performance.
problem Uniform-reward RLHF fails to capture diverse preferences, leading to unfairness.
method SharedRep-RLHF learns shared traits among various groups, improving fairness and performance.
result SharedRep-RLHF outperforms MaxMin-RLHF by up to 20% in win rate.
New suboptimal algorithm for best subset selection in high-dimensional data.
problem Nonconvex and computationally challenging best subset selection in linear regression.
method Introducing a new suboptimal algorithm and comparing it with other popular methods.
result The new procedure is a competitive suboptimal algorithm for high-dimensional data.
The maximum a posteriori (MAP) configuration of binary variable models with submodular graph-structured energy functions can be found efficiently and exactly by graph cuts. Max-product belief propagation (MP) has been shown to be suboptimal on this class of energy functions by a canonical counterexample where MP conver…
The rollout of new versions of a feature in modern applications is a manual multi-stage process, as the feature is released to ever larger groups of users, while its performance is carefully monitored. This kind of A/B testing is ubiquitous, but suboptimal, as the monitoring requires heavy human intervention, is not gu…
The Lasso is suboptimal in sparse linear regression due to design matrix constraints.
problem The suboptimality of the Lasso estimator in sparse linear regression.
method Characterization of optimal estimation rates and analysis of the Lasso estimator.
result The Lasso is provably minimax rate-suboptimal when the minimum singular value is small.
The broad set of deep generative models (DGMs) has achieved remarkable advances. However, it is often difficult to incorporate rich structured domain knowledge with the end-to-end DGMs. Posterior regularization (PR) offers a principled framework to impose structured constraints on probabilistic models, but has limited …
RL optimizes resource allocation in MG by balancing experience and exploration.
problem Optimal resource allocation in competitive scenarios.
method Introduced RL to MG, allowing dynamic strategy adjustment based on experience and expected rewards.
result Achieves optimal resource coordination by balancing exploitation and exploration.
Paper explores limits of imitation learning in MDPs, setting new suboptimality bounds.
problem Understanding the statistical limits of imitation learning in MDPs.
method Analyzes minimax statistical limits in two settings: pre-interaction and interaction.
result Establishes suboptimality bounds for imitation learning in MDPs, showing improvements with knowledge of transition.
New bounds on adaptivity cost in stochastic optimization.
problem Understanding the cost of changing strategies in stochastic optimization.
method Proving impossibility results for adaptivity in non-smooth stochastic convex optimization.
result Lower bounds on the price of adaptivity for different levels of uncertainty.
Practical model building processes are often time-consuming because many different models must be trained and validated. In this paper, we introduce a novel algorithm that can be used for computing the lower and the upper bounds of model validation errors without actually training the model itself. A key idea behind ou…
Enhances reinforcement learning from sparse data.
problem Limited data for offline reinforcement learning.
method Trajectory-based data augmentation.
result Improves reinforcement learning performance.
New algorithm reduces suboptimality in imitation learning to nearly optimal levels.
problem Statistical limits of imitation learning in MDPs with known transitions.
method Mimic-MD algorithm and reduction to value estimation problem.
result Upper bound of O(∣S∣H3/2/N) for suboptimality, with efficient computation. Least Squares Estimators are suboptimal for 5D convex functions.
problem Suboptimality of Least Squares Estimators in estimating multidimensional convex functions.
method Analysis of natural subclasses of convex functions in random and fixed design settings.
result Risk of LSE is n−2/d while minimax risk is n−4/(d+4) for d≥5. New study shows exponential lower bound for RL even with constant suboptimality gap.
problem Can RL be sample-efficient with a constant suboptimality gap?
method Analyzes reinforcement learning in the online setting with a linearly realizable optimal Q-function.
result An exponential sample complexity lower bound still holds even with a constant suboptimality gap.
Training a source model optimally for its own task is suboptimal for downstream transfer.
problem The optimality of a source model for its own task hinders downstream transfer performance.
method Analyzes L2-SP ridge regression, characterizes transfer-optimal source penalty, and identifies alignment-dependent effects.
result Transfer benefits from stronger source regularization when aligned imperfectly, and from weaker regularization when aligned perfectly.
A new measure helps compute suboptimality in entropy-regularized methods.
problem Computing suboptimality in entropy-regularized variational objectives when unnormalised densities are unavailable.
method Introduced 'kernel gradient discrepancy' (KGD) to compute suboptimality explicitly.
result KGD characterizes kernel Stein discrepancy (KSD) in the standard Bayesian context and measures variational gradient size.
Understanding the loss surface of neural networks is essential for the design of models with predictable performance and their success in applications. Experimental results suggest that sufficiently deep and wide neural networks are not negatively impacted by suboptimal local minima. Despite recent progress, the reason…
Posterior inference in directed graphical models is commonly done using a probabilistic encoder (a.k.a inference model) conditioned on the input. Often this inference model is trained jointly with the probabilistic decoder (a.k.a generator model). If probabilistic encoder encounters complexities during training (e.g. s…
The PC algorithm is a popular method for learning the structure of Gaussian Bayesian networks. It carries out statistical tests to determine absent edges in the network. It is hence governed by two parameters: (i) The type of test, and (ii) its significance level. These parameters are usually set to values recommended …
Paper explores PG for MCR, finding suboptimal policies but providing bounds.
problem Optimizing coherent risk in MDPs using PG.
method Developed practical PG implementation using state distribution reweighting.
result PG can learn risk-sensitive policies but suboptimality is common.
Online Multi-Object Tracking (MOT) from videos is a challenging computer vision task which has been extensively studied for decades. Most of the existing MOT algorithms are based on the Tracking-by-Detection (TBD) paradigm combined with popular machine learning approaches which largely reduce the human effort to tune a…
Negative momentum accelerates convergence in minimax games but at a suboptimal rate.
problem The convergence rate of negative momentum in minimax games is suboptimal.
method Extending variational inequality formulation, connecting momentum method with Chebyshev polynomials.
result Negative momentum accelerates convergence locally but at a suboptimal rate.
In this paper, we propose and study opportunistic bandits - a new variant of bandits where the regret of pulling a suboptimal arm varies under different environmental conditions, such as network load or produce price. When the load/price is low, so is the cost/regret of pulling a suboptimal arm (e.g., trying a suboptim…
MTAdam optimizes multiple loss terms in neural models, balancing gradients dynamically.
problem Balancing multiple loss terms in neural model training is challenging and computationally demanding.
method Generalized Adam algorithm that computes separate derivatives and balances gradients across layers dynamically.
result Training with MTAdam leads to faster recovery from suboptimal initial loss weighting and matches conventional training outcomes.
RL-LOW algorithm achieves exponential simple regret in offline RLHF with pairwise comparisons.
problem Offline reinforcement learning from human feedback with pairwise comparisons.
method Proposes RL-LOW algorithm to minimize simple regret with exponential convergence.
result Achieves an exponential form of simple regret of \(\exp ( - Ω(n/H) )\).
The paper tackles fast rates in batch active learning with pool-based data.
problem Reduced adaptivity in batch active learning leads to suboptimal results.
method Proposes a stage-wise greedy algorithm that balances informativeness and diversity.
result The algorithm's excess risk matches minimax rates in standard statistical learning settings.
Interpolating estimators in nonparametric regression become suboptimal under adversarial attacks.
problem Adversarial robustness of interpolating estimators in nonparametric regression.
method Investigation of adversarial robustness of interpolating estimators in a nonparametric regression framework.
result Interpolating estimators must be suboptimal even under a subtle future X-attack. New algorithms tackle robust RL with linear models, revealing unique challenges.
problem Distributionally robust offline RL with uncertainty in dynamics.
method Proposes minimax optimal and computationally efficient algorithms using novel function approximation mechanisms.
result Function approximation in robust offline RL is distinct and harder than in standard offline RL.
Using a fast numerical technique, we investigate a large database of investor suboptimal non-exercise of short maturity American call options on dividend-paying stocks listed on the Dow Jones. The correct modelling of the discrete dividend is essential for a correct calculation of the early exercise boundary as confirm…