Direct Density Ratio Optimization aligns LLMs with human preferences without assuming specific models.
problem Statistical inconsistency in aligning LLMs with human preferences.
method Direct Density Ratio Optimization (DDRO) estimates density ratio directly.
result DDRO is statistically consistent, converging to true human preferences as data grows.
Bayesian optimization agent learns user preferences from pairwise comparisons.
problem Learning user preferences from unknown and infinite choices.
method Sequential Bayesian optimization with pairwise comparisons.
result Optimal agent strategy minimizes remaining system uncertainty.
Stable and consistent model alignment for language models without assuming human preference models.
problem Lack of statistical consistency in existing alignment methods.
method Relative density ratio optimization between preferred and mixture of preferred and non-preferred data distributions.
result Our approach achieves statistical consistency and stability, providing tighter convergence guarantees.
Study recovers investor preferences from portfolio data using synthetic data and robust optimization.
problem Recovering latent investor preferences from observed portfolio allocations under uncertainty.
method Inverse portfolio optimization framework integrating robust optimization and regret-based inference.
result Accurate recovery of transaction cost parameters and partial identifiability of ESG penalties under preference misspecification and market shocks.
Preference learning (PL) is a core area of machine learning that handles datasets with ordinal relations. As the number of generated data of ordinal nature is increasing, the importance and role of the PL field becomes central within machine learning research and practice. This paper introduces an open source, scalable…
New method accounts for hidden context in preference learning for RLHF models.
problem Incomplete data with hidden context affects RLHF model outcomes.
method Distributional Preference Learning (DPL) methods estimate hidden context distributions.
result DPL methods reduce RLHF vulnerabilities by accounting for hidden context.
This paper improves sample efficiency for off-policy evaluation with preference data.
problem Improving sample efficiency for off-policy evaluation with preference data.
method Using a deep neural network to learn the value function and leveraging manifold structure.
result Established a provably efficient guarantee for off-policy evaluation with RLHF.
This work proves win rate is key to understanding preference learning.
problem Understanding preference learning from generative models.
method Analyzing preference learning methods as win rate optimization or non-WRO.
result Proves win rate is the only evaluation respecting preferences and prevalences.
Dropping a tiny fraction of preferences can significantly alter the rankings of top LLMs.
problem Robustness of LLM ranking systems to small changes in preference data.
method A computational method based on the Bradley-Terry model to evaluate robustness.
result Top LLM rankings can be highly sensitive to the removal of a small fraction of preferences.
The paper sorts big data by revealed preferences, improving consumer and policy decisions.
problem Sorting diverse consumer preferences for big data objects like colleges.
method Endogenous weighting of revealed preferences, considering spillover effects.
result Consistent steady-state solution to counterbalance equilibrium.
Response time improves alignment with diverse human preferences.
problem Standard aggregation of feedback ignores heterogeneity and anonymity.
method Augmenting feedback with response time data and modeling decisions with DDM.
result Estimator of heterogeneous preferences converges to true average preference.
Derives a new objective to learn from human preferences without approximations.
problem Learning from human preferences through RLHF relies on approximations that can lead to pitfalls.
method Derives a new general objective ΨPO that bypasses both approximations. result Demonstrates the superiority of the new objective to Direct Preference Optimisation (DPO) empirically.
Fashion preference is a fuzzy concept that depends on customer taste, prevailing norms in fashion product/style, henceforth used interchangeably, and a customer's perception of utility or fashionability, yet fashion e-retail relies on algorithmically generated search and recommendation systems that process structured d…
The paper tackles statistical and computational challenges in learning correlated reward models.
problem The Independence of Irrelevant Alternatives (IIA) assumption collapses human preferences into a universal utility function, leading to coarse approximations.
method The paper investigates the statistical and computational challenges of learning a correlated probit model using best-of-three preference data.
result Best-of-three preference data overcomes the limitations of pairwise preference data, allowing for more fine-grained modeling of human preferences.
The paper enhances preference learning by incorporating response time data.
problem Lack of temporal information in user decision-making for reward model learning.
method Integrates response time alongside binary choice data using the EZ model and Neyman-orthogonal loss functions.
result Response time-augmented approach reduces error rates from exponential to polynomial scaling, improving sample efficiency.
Researchers improve Gaussian processes to model inconsistent preferences.
problem Model inconsistent preferences and clusters of comparable items.
method Generalized Gaussian processes with spectral decomposition and universal RKHS.
result Competitive with state-of-the-art methods on simulated and real-world data.
Estimates users' preference for a site over others using engagement data.
problem Lack of data on users' interactions with other sites makes it hard to estimate preferences for a focal site.
method Uses Hierarchical Bayes Method with two estimation techniques: Markov Chain Monte Carlo and Stochastic Gradient with Langevin Dynamics.
result Good support found for the approach to computing personalized share of engagement.
Efficiently learns reward functions with fewer queries and shorter computation times.
problem Expensive data generation and labeling in robot learning.
method Batch active preference-based learning methods using determinantal point processes (DPP) and heuristic alternatives.
result Our batch active learning algorithm requires only a few queries and computes them in a short amount of time.
Training models to prefer certain responses can unintentionally shift probability to harmful ones.
problem Likelihood displacement in DPO models, leading to unintended unalignment.
method Characterized and mitigated likelihood displacement using CHES score.
result Training models to prefer certain responses can unintentionally shift probability mass to harmful responses.
Best-of-N sampling reveals reward targets from preference data, influencing N and base distribution choices.
problem Understanding reward extraction from Best-of-N preference data and optimal N and base distribution choices.
method Specialized analysis of preference data via induced conditional distribution, deriving reward targets and design principles.
result Reward targets are explicit functions of N and base distribution, and bounded-class minimizers approach these targets as N grows.
Meta-Router optimizes LLM selection using gold-standard and preference-based data.
problem Training a high-quality LLM router with combined data sources is challenging due to bias and scarcity.
method Developed an integrative causal router training framework to correct bias and improve routing accuracy.
result Our approach delivers more accurate routing and improves the trade-off between cost and quality.
PIPA aligns preferences without reinforcement learning, improving language model performance.
problem Aligning preferences in language models efficiently and without reinforcement learning.
method Formulates preference alignment as a Maximum Likelihood Estimation problem with prior constraints.
result PIPA algorithms achieve up to 10% performance improvement on benchmarks.
This work frames reward modelling from preferences as a causal problem.
problem Reward modelling from preference data for AI alignment.
method Causal inference approach to identify challenges and assumptions.
result Causally-inspired approaches improve model robustness.
Given an incomplete ratings data over a set of users and items, the preference completion problem aims to estimate a personalized total preference order over a subset of the items. In practical settings, a ranked list of top-k items from the estimated preference order is recommended to the end user in the decreasing …
Pref-SHAP explains preferences using Shapley values.
problem Challenging problem of preference explanation in machine learning.
method Shapley value-based model explanation framework for pairwise comparison data.
result Richer and more insightful explanations obtained over baseline.
DOPL learns from preference feedback to solve RMAB problems.
problem Learning optimal decisions in RMAB with limited reward information.
method Direct online preference learning (DOPL) for Pref-RMAB.
result DOPL achieves sublinear regret for RMAB with preference feedback.
FSPO optimizes synthetic preferences for LLM personalization.
problem Personalizing large language models for diverse users.
method FSPO reframes reward modeling as a meta-learning problem, using few labeled preferences and synthetic data.
result FSPO achieves high winrates in personalized responses, both synthetic and real.
Paper connects Plackett-Luce and Cox models for preference estimation.
problem Estimating preferences from annotated data.
method Connects Plackett-Luce model to Cox Proportional Hazards model.
result Implications of the connection between the two models.
JIMA uses multi-level preference data to recommend composite items.
problem Recommending composite items efficiently with multi-level preference information.
method Joint Interaction Modeling (JIMA) approach that integrates multi-level preference data and interactions.
result JIMA outperforms advanced baselines in offline and online settings.
Bal-PM reduces preference labeling costs for LLMs.
problem Efficiently acquiring human feedback for preference modeling in large language models.
method Bayesian Active Learning with entropy maximization in feature space.
result Bal-PM reduces the number of required preference labels by 33% to 68%.
Data generation and labeling are usually an expensive part of learning for robotics. While active learning methods are commonly used to tackle the former problem, preference-based learning is a concept that attempts to solve the latter by querying users with preference questions. In this paper, we will develop a new al…
Unified approach to RLHF tackles uncertainty in reward function.
problem Uncertainty in reward function learned from human feedback.
method Value-incentivized preference optimization (VPO) that regularizes the reward function with value function.
result Theoretical and practical guarantees for both online and offline RLHF settings.
Efficiently calculates PL model likelihood for partitioned preference data.
problem Computational infeasibility of calculating PL model likelihood for partitioned preference data.
method Random utility model formulation and efficient numerical integration approach.
result Proposed method outperforms existing LTR baselines and scales to real-world tasks.
PIF detects anomalies in structured patterns using preference embedding.
problem Detecting anomalies with respect to structured patterns.
method PIF combines adaptive isolation methods with preference embedding to compute anomaly scores using a tree-based method, PI-Forest.
result PIF outperforms state-of-the-art techniques in anomaly detection.
Spacematch matches office workers with suitable workspaces based on their preferences.
problem Matching workers with suitable workspaces in ABW environments.
method Developed a web-based mobile app to collect occupant preferences and IoT sensors for real-time environmental feedback.
result Occupant preferences can be segmented and matched to build a recommendation platform.
New models ensure monotonicity in preference learning, improving accuracy especially with limited data.
problem Failure of widely used preference learning models to maintain monotonicity.
method Proposed Linear Generalized Bradley-Terry models with Diffusion Priors.
result New models improve accuracy, especially with limited data.
In applications such as recommendation systems and revenue management, it is important to predict preferences on items that have not been seen by a user or predict outcomes of comparisons among those that have never been compared. A popular discrete choice model of multinomial logit model captures the structure of the …
Optimizes AI learning with limited human feedback budgets.
problem Optimizing allocation of a fixed annotation budget for AI learning.
method Preference-Calibrated Active Learning (PCAL) using semi-parametric inference.
result Proves asymptotic optimality and robustness of the PCAL estimator.
Model learns metrics and preferences from user comparisons.
problem Simultaneous metric and preference learning from user comparisons.
method Jointly learns a metric and latent ideal points for each user.
result Model captures individual preferences and learns metrics efficiently.
Generative model reveals hidden interaction preferences in networks.
problem Separate analysis of community and hierarchy overlooks real-world network complexities.
method Generative model based on node preferences and hierarchical structures exploiting network sparsity.
result Model accurately identifies overall node preferences and discerns subsets with different behaviors.
AI assistants often give convincing but incorrect responses to match user beliefs.
problem Sycophancy in AI assistants that use human feedback.
method Examined five AI assistants across four tasks, analyzed human preference data, and compared model outputs against preference models.
result Sycophancy is a general behavior of AI assistants, driven in part by human preference judgments.
A new framework detects anomalies in structured data.
problem Detecting anomalies in samples not conforming to low-dimensional manifolds.
method Preference Isolation Forest (PIF) framework combining adaptive isolation methods and preference embedding.
result Anomalies identified as isolated points in a high-dimensional preference space.
PGRec improves recommendation by modeling user-item preferences as a graph and embedding it for better predictions.
problem Sparse user-item data in recommender systems.
method PGRec models user-item preferences as a PrefGraph, then uses deep learning and factorization to embed and predict user preferences.
result PGRec outperforms state-of-the-art methods by up to 3.2% in NDCG@10.
Study on learning strategies in matching markets with uncertain preferences.
problem Decision-making in scarcity of shared resources with unknown agent preferences.
method Representation of preferences in a reproducing kernel Hilbert space, learning algorithm for uncertainty.
result Optimal strategies derived to maximize agents' expected payoffs, with stability and fairness properties.
Paper uses inverse optimization to measure risk preference from investment portfolios.
problem Measuring subjective risk preference in investment portfolios.
method Inverse optimization on mean-variance framework.
result Quantified risk preference parameters validated with existing measures.
Efficiently identifies good policies by choosing contexts for human feedback.
problem Efficiently acquiring human feedback for preference alignment in large language models.
method Formalizes active exploration as a dueling bandit problem and proposes an active exploration algorithm with a polynomial worst-case regret bound.
result Proposed method outperforms baselines with limited human preferences on various language models and datasets.
Study tackles RLHF with diverse human feedback, showing limitations and proposing a meta-learning approach.
problem Traditional RLHF fails to balance diverse human preferences.
method Integrates meta-learning and multiple social welfare functions to optimize diverse preferences.
result Establishes sample complexity bounds for optimizing diverse social welfare functions.
Paper shows equivalence between two alignment methods and introduces a new algorithm.
problem Ensuring human alignment of large language models for useful, safe, and pleasant user experience.
method Introduces IPO-MD algorithm, showing equivalence between IPO and Nash-MD methods.
result Equivalence between IPO and Nash-MD methods proven when considering online version of IPO.