DPA aligns LLMs with multi-objective rewards for diverse user preferences.
problem Fine-grained control over LLMs for diverse user needs.
method Integrates multi-objective reward modeling and directional preference control.
result DPA offers better performance trade-offs and intuitive user control over LLM generation.
Improved DPO framework penalizes preference uncertainty to avoid overoptimization.
problem Aligning LLMs to human preferences is challenging due to varied, context-dependent, and ambiguous preferences.
method Developed a pessimistic framework for DPO by introducing preference uncertainty penalization schemes.
result Improved overall performance and better completions on high-uncertainty responses compared to vanilla DPO.
Active learning framework for optimizing human preferences in reinforcement learning.
problem Selecting most informative feedback for training models of human preferences.
method Proposes an active learning framework to collect preferential feedback online or offline.
result Errors in DPO logit estimates diminish with more feedback.
Enhances math problem-solving models with multi-turn preference learning.
problem Improving mathematical problem-solving capabilities of large language models.
method Introduces a multi-turn direct preference learning framework for tool-integrated mathematical reasoning tasks.
result Significant performance improvements in model accuracy on math datasets.
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.
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.
A-GPS learns to generate Pareto sets efficiently with user preferences.
problem Online discrete multi-objective optimization with user preferences.
method Generative model with class probability estimator (CPE) for non-dominance and preference alignment.
result Amortized generative model for efficient Pareto set approximation.
Paper investigates monotonicity issues in AI preference learning.
problem AI models may violate monotonicity when learning preferences.
method Investigates root causes of non-monotonicity in comparison-based preference learning.
result Proves local pairwise monotonicity under mild assumptions.
This paper introduces f-DPO, a generalized approach to Direct Preference Optimization using diverse divergence constraints.
problem Aligning large language models with human preferences while mitigating safety risks.
method Incorporates diverse divergence constraints to simplify the relationship between reward and optimal policy, eliminating the need for estimating the normalizing constant.
result Optimizes LLMs to align with human preferences more efficiently and under a broader set of divergence constraints.
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.
Optimizes explanations for better listener understanding.
problem Insufficient consideration of listener preferences in concept-based explanations.
method Iterative training procedure based on direct preference optimization.
result Pragmatic explanations improve both model accuracy and user understanding.
Optimizes molecular generation for chemist preferences.
problem Models lack inherent preferences for chemist-desired structures.
method Fine-tuning with Direct Preference Optimization.
result Approach is simple, efficient, and highly effective.
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.
New algorithm for learning preferences in decentralized matching markets reduces regret to logarithmic levels.
problem Learning preferences in decentralized matching markets without direct communication.
method Introduces a new algorithm for two-sided matching markets with competition.
result The algorithm achieves logarithmic stable regret in shared preferences and quadratic regret in general preferences.
New methods improve LLM preference optimization by intelligently weighting multiple reference models.
problem Improving LLM preference optimization with multiple reference models.
method Introducing four new weighting strategies for multiple-reference preference optimization.
result All four new weighting strategies outperform current methods on preference accuracy.
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.
A particular Finsler-metric proposed in [1,2] and describing a geometry with a preferred null direction is characterized here as belonging to a subclass contained in a larger class of Finsler-metrics with one or more preferred directions (null, space- or timelike). The metrics are classified according to their group of…
MallowsPO enhances LLM fine-tuning with a dispersion index of human preferences.
problem Lack of diversity in human preferences in DPO.
method Developed a dispersion index based on Mallows' theory to characterize preference diversity.
result Demonstrated improved performance in various tasks using the dispersion index.
Paper develops methods to optimize policies directly from human feedback without reward inference.
problem Challenges in RLHF, including reward model overfitting and distribution shift.
method Develops two algorithms for RLHF without reward inference, using zeroth-order gradient approximators.
result Establishes polynomial convergence rates and outperforms existing methods in numerical experiments.
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.
Efficiently poisons offline RLHF models by flipping preference labels.
problem Vulnerability of offline RLHF models to preference label flipping attacks.
method Developed two attack methods: BAL-A and BMP-A, solving a structured binary sparse approximation problem.
result Demonstrated that flipping one preference label induces a parameter-independent shift in the DPO gradient, enabling structured binary sparse approximation.
New method improves consistency in preference learning for neural networks.
problem Inconsistent surrogate losses in preference learning for neural networks.
method Formulated a margin-shifted ranking framework and introduced Structure-Aware H-consistency. result Proved superior consistency guarantees for capacity-bounded models using heavy-tailed surrogates.
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.
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.
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.
Policy gradient algorithms typically combine discounted future rewards with an estimated value function, to compute the direction and magnitude of parameter updates. However, for most Reinforcement Learning tasks, humans can provide additional insight to constrain the policy learning. We introduce a general method to i…
Due to the drastic increase of mobile traffic, wireless caching is proposed to serve repeated requests for content download. To determine the caching scheme for decentralized caching networks, the content preference learning problem based on mobility prediction is studied. We first formulate preference prediction as a …
New method detects inconsistencies in AHP matrices using triadic preference reversals.
problem Challenges in assessing consistency in AHP pairwise comparison matrices.
method Triadic preference reversals to detect inconsistencies between pairs of elements.
result 97% accuracy in detecting inconsistencies, significantly surpassing traditional methods.
Proposes a method for ranking items across multiple aspects based on user feedback.
problem No principled solution exists for generating multiple item rankings over different aspects.
method Developed a directional multi-aspect ranking criterion using probabilistic multivariate tensor factorization.
result Demonstrated effectiveness of the proposed method through comprehensive experiments on real datasets.
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.
New method for efficient online exploration in RLHF reduces regret.
problem Efficiently collecting new preference data in RLHF to refine reward model and policy.
method Proposes a new exploration scheme that directs preference queries toward reducing uncertainty in reward differences most relevant to policy improvement.
result Establishes regret bounds of order T(β+1)/(β+2) for online RLHF, with polynomial scaling in all model parameters. Study creates web interface to elicit user-preferred metrics.
problem Eliciting classification metrics that align with user preferences.
method Developed a web-based interface and conducted a user study.
result Users preferred metrics that align with their task and context.
Paper addresses RLHF alignment challenges with novel algorithms.
problem Challenges in RLHF alignment, especially in strategic exploration.
method Develops a reverse-KL regularized contextual bandit formulation and proposes efficient algorithms with theoretical guarantees.
result Proposed methods significantly outperform existing RLHF algorithms in real-world experiments.
Mitigates overoptimization in RLHF by reformulating SFT loss as a preference optimization loss.
problem Overoptimization in RLHF where reward model misguides generative model.
method Proposes a theoretical algorithm that minimizes maximum likelihood estimation and reward penalty, reformulates as simple objective combining preference optimization and supervised learning losses.
result Improved performance of RPO compared to DPO baselines in aligning LLMs.
Paper proposes efficient RLHF methods for LLMs using active queries.
problem Efficiently aligning LLMs with human preferences using RLHF.
method Formalizes RLHF as a dueling bandit problem, introduces APPO and ADPO algorithms.
result ADPO achieves similar performance to state-of-the-art methods with fewer queries.
A new algorithm RTO integrates DPO and PPO for RLHF, improving performance.
problem Learning from sparse, sentence-level rewards in RLHF.
method Modeling RLHF as MDP, introducing RTO that learns token-wise reward function.
result RTO outperforms PPO on benchmarks by 7.5 points on AlpacaEval 2 and 4.1 points on Arena-Hard.
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.
ADPO optimizes relative advantage in reinforcement learning from human feedback.
problem Optimizing policy alignment in reinforcement learning from human preferences.
method ADPO explicitly parameterizes the optimal structure through anchored logits, decoupling response quality from prior popularity.
result Empirically, ADPO achieves state-of-the-art performance on reasoning tasks, outperforming GRPO by 30.9 percent.
Unified algorithm tackles various RL goals like reward-free and preference-based learning.
problem Unified approach to multiple RL learning goals.
method Decision-Estimation Coefficient (DEC) framework.
result Unified algorithm handles various learning goals with a single framework.
EXPO framework eliminates need for reward model, achieving better optimization.
problem Optimizing large language model responses without a separate reward model.
method Introduces EXPO framework that avoids reparameterization, directly optimizing preferences.
result Demonstrates better regularization and intuitive interpolation behaviors.
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.
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.
ZSPO optimizes RL from unknown link functions using human feedback.
problem Designing RLHF algorithms for unknown link functions.
method Zero-order policy optimization with human preference feedback.
result ZSPO converges to a stationary policy with a polynomial rate.
Many currently deployed Reinforcement Learning agents work in an environment shared with humans, be them co-workers, users or clients. It is desirable that these agents adjust to people's preferences, learn faster thanks to their help, and act safely around them. We argue that most current approaches that learn from hu…
Investigates how diversification preferences relate to risk attitudes.
problem Connecting diversification preferences to risk attitudes.
method Analyzes diversification preferences for various pairs of risks under different conditions.
result Diversification preferences for certain pairs of risks imply specific levels of risk aversion.
This work simplifies data valuation for LLMs using Shapley value computation.
problem How to fairly distribute benefits from training superior LLMs with multiple data owners' resources.
method We leverage the specific mathematical structure of DPO to enable scalable Shapley value computation for LLMs.
result We demonstrate that Shapley value computation for LLMs trained with DPO is significantly simplified.
The paper presents a new method to represent directed graphs using pseudo-Riemannian manifolds.
problem Representing directed graphs in a compact and meaningful way.
method Combines pseudo-Riemannian metric structure, non-trivial global topology, and a unique likelihood function.
result Low-dimensional cylindrical Minkowski and anti-de Sitter spacetimes produce equal or better graph representations than curved Riemannian manifolds.
A new method for RLHF using proximal point Nash learning.
problem Capturing real human preferences in RLHF.
method Proximal point Nash learning, embedding self-play updates into a proximal point framework.
result High-probability last-iterate convergence for the combined method.