Existing pruning methods for large language models (LLMs) focus on achieving high compression rates while maintaining model performance. Although these methods have demonstrated satisfactory performance in handling a single user's compression request, their processing time increases linearly with the number of requests, making them inefficient for real-world scenarios with multiple simultaneous requests. To address this limitation, we propose a Univeral Model for Customized Compression (UniCuCo) for LLMs, which introduces a StratNet that learns to map arbitrary requests to their optimal pruning strategy. The challenge in training StratNet lies in the high computational cost of evaluating pruning strategies and the non-differentiable nature of the pruning process, which hinders gradient backpropagation for StratNet updates. To overcome these challenges, we leverage a Gaussian process to approximate the evaluation process. Since the gradient of the Gaussian process is computable, we can use it to approximate the gradient of the non-differentiable pruning process, thereby enabling StratNet updates. Experimental results show that UniCuCo is 28 times faster than baselines in processing 64 requests, while maintaining comparable accuracy to baselines.

This paper presents Multi-Objective Reinforcement Learning from AI Feedback (MORLAIF), a novel approach to improving the alignment and performance of language models trained using reinforcement learning from AI feedback (RLAIF). In contrast to standard approaches that train a single preference model to represent all human preferences, MORLAIF decomposes this task into multiple simpler principles, such as toxicity, factuality, and sycophancy. Separate preference models are trained for each principle using feedback from GPT-3.5-Turbo. These preference model scores are then combined using different scalarization functions to provide a reward signal for Proximal Policy Optimization (PPO) training of the target language model. Our experiments indicate that MORLAIF outperforms the standard RLAIF baselines and that MORLAIF can be used to align larger language models using smaller ones. Surprisingly, the choice of scalarization function does not appear to significantly impact the results. Recent advancements in large language models (LLMs) have led to remarkable performance across a wide range of natural language tasks (Bommasani et al. (2021); Brown et al. (2020)). However, ensuring that these models behave in alignment with human values and preferences remains a significant challenge (Kenton et al. (2021)). Reinforcement learning from human feedback (RLHF) has emerged as a promising approach to address this issue by training models to optimize for humanspecified reward functions (Christiano et al. (2017); Stiennon et al. (2020)), but many issues with RLHF have been identified, such as the limited ability of humans to evaluate responses and reward hacking of the preference model (Casper et al. (2023)). In a standard RLHF setup, a preference model is trained on human comparisons of model outputs to represent human preferences.

Multi-objective reinforcement learning (MORL) extends traditional RL by seeking policies making different compromises among conflicting objectives. The recent surge of interest in MORL has led to diverse studies and solving methods, often drawing from existing knowledge in multi-objective optimization based on decomposition (MOO/D). Yet, a clear categorization based on both RL and MOO/D is lacking in the existing literature. Consequently, MORL researchers face difficulties when trying to classify contributions within a broader context due to the absence of a standardized taxonomy. To tackle such an issue, this paper introduces multi-objective reinforcement learning based on decomposition (MORL/D), a novel methodology bridging the literature of RL and MOO. A comprehensive taxonomy for MORL/D is presented, providing a structured foundation for categorizing existing and potential MORL works. The introduced taxonomy is then used to scrutinize MORL research, enhancing clarity and conciseness through well-defined categorization. Moreover, a flexible framework derived from the taxonomy is introduced. This framework accommodates diverse instantiations using tools from both RL and MOO/D. Its versatility is demonstrated by implementing it in different configurations and assessing it on contrasting benchmark problems. Results indicate MORL/D instantiations achieve comparable performance to current state-of-the-art approaches on the studied problems. By presenting the taxonomy and framework, this paper offers a comprehensive perspective and a unified vocabulary for MORL. This not only facilitates the identification of algorithmic contributions but also lays the groundwork for novel research avenues in MORL.

The goal of multi-objective optimization is to identify a collection of points which describe the best possible trade-offs between the multiple objectives. In order to solve this vector-valued optimization problem, practitioners often appeal to the use of scalarization functions in order to transform the multi-objective problem into a collection of single-objective problems. This set of scalarized problems can then be solved using traditional single-objective optimization techniques. In this work, we formalise this convention into a general mathematical framework. We show how this strategy effectively recasts the original multi-objective optimization problem into a single-objective optimization problem defined over sets. An appropriate class of objective functions for this new problem is the R2 utility function, which is defined as a weighted integral over the scalarized optimization problems. We show that this utility function is a monotone and submodular set function, which can be optimised effectively using greedy optimization algorithms. We analyse the performance of these greedy algorithms both theoretically and empirically. Our analysis largely focusses on Bayesian optimization, which is a popular probabilistic framework for black-box optimization.

Machine learning (ML) methods offer a wide range of configurable hyperparameters that have a significant influence on their performance. While accuracy is a commonly used performance objective, in many settings, it is not sufficient. Optimizing the ML models with respect to multiple objectives such as accuracy, confidence, fairness, calibration, privacy, latency, and memory consumption is becoming crucial. To that end, hyperparameter optimization, the approach to systematically optimize the hyperparameters, which is already challenging for a single objective, is even more challenging for multiple objectives. In addition, the differences in objective scales, the failures, and the presence of outlier values in objectives make the problem even harder. We propose a multi-objective Bayesian optimization (MoBO) algorithm that addresses these problems through uniform objective normalization and randomized weights in scalarization. We increase the efficiency of our approach by imposing constraints on the objective to avoid exploring unnecessary configurations (e.g., insufficient accuracy). Finally, we leverage an approach to parallelize the MoBO which results in a 5x speed-up when using 16x more workers.

Single-objective black box optimization (also known as zeroth-order optimization) is the process of minimizing a scalar objective $f(x)$, given evaluations at adaptively chosen inputs $x$. In this paper, we consider multi-objective optimization, where $f(x)$ outputs a vector of possibly competing objectives and the goal is to converge to the Pareto frontier. Quantitatively, we wish to maximize the standard hypervolume indicator metric, which measures the dominated hypervolume of the entire set of chosen inputs. In this paper, we introduce a novel scalarization function, which we term the hypervolume scalarization, and show that drawing random scalarizations from an appropriately chosen distribution can be used to efficiently approximate the hypervolume indicator metric. We utilize this connection to show that Bayesian optimization with our scalarization via common acquisition functions, such as Thompson Sampling or Upper Confidence Bound, provably converges to the whole Pareto frontier by deriving tight hypervolume regret bounds on the order of $\widetilde{O}(\sqrt{T})$. Furthermore, we highlight the general utility of our scalarization framework by showing that any provably convergent single-objective optimization process can be effortlessly converted to a multi-objective optimization process with provable convergence guarantees.

A BSTRACT In the multi-objective reinforcement learning (MORL) paradigm, the relative importance of each environment objective is often unknown prior to training, so agents must learn to specialize their behavior to optimize different combinations of environment objectives that are specified post-training. These are typically linear combinations, so the agent is effectively parameterized by a weight vector that describes how to balance competing environment objectives. However, many real world behaviors require nonlinear combinations of objectives. Additionally, the conversion between desired behavior and weightings is often unclear. In this work, we explore the use of a language based on propositional logic with quantitative semantics-in place of weight vectors-for specifying nonlinear behaviors in an interpretable way. We use a recurrent encoder to encode logical combinations of objectives, and train a MORL agent to generalize over these encodings. We test our agent in several grid worlds with various objectives and show that our agent can generalize to many never-before-seen specifications with performance comparable to single policy baseline agents. We also demonstrate our agent's ability to generate meaningful policies when presented with novel specifications and quickly specialize to novel specifications. 1 I NTRODUCTION Reinforcement Learning (RL) is a method for learning behavior policies by maximizing expected reward through interactions with an environment. RL has grown in popularity as RL agents have excelled at increasingly complex tasks, including board games (Silver et al., 2016), video games (Mnih et al., 2015), robotic control (Haarnoja et al., 2018), and other high dimensional, complex tasks.

Abstract-- Multi-objective reinforcement learning (MORL) is the generalization of standard reinforcement learning (RL) approaches to solve sequential decision making problems that consist of several, possibly conflicting, objectives. Generally, in such formulations, there is no single optimal policy which optimizes all the objectives simultaneously, and instead, a number of policies has to be found, each optimizing a preference of the objectives. In this paper, we introduce a novel MORL approach by training a meta-policy, a policy simultaneously trained with multiple tasks sampled from a task distribution, for a number of randomly sampled Markov decision processes (MDPs). In other words, the MORL is framed as a meta-learning problem, with the task distribution given by a distribution over the preferences. We demonstrate that such a formulation results in a better approximation of the Pareto optimal solutions, in terms of both the optimality and the computational efficiency. We evaluated our method on obtaining Pareto optimal policies using a number of continuous control problems with high degrees of freedom. I. INTRODUCTION Reinforcement learning (RL) is a framework to train an agent to acquire a behavior by reinforcing actions that maximize a notion of task-relevant future rewards. A reward function, i.e., the function that assigns a reward value to every action-decision made by the agent, is designed to guide the training to implement the behavior.

Sequential decision-making problems with multiple objectives arise naturally in practice and pose unique challenges for research in decision-theoretic planning and learning, which has largely focused on single-objective settings. This article surveys algorithms designed for sequential decision-making problems with multiple objectives. Though there is a growing body of literature on this subject, little of it makes explicit under what circumstances special methods are needed to solve multi-objective problems. Therefore, we identify three distinct scenarios in which converting such a problem to a single-objective one is impossible, infeasible, or undesirable. Furthermore, we propose a taxonomy that classifies multi-objective methods according to the applicable scenario, the nature of the scalarization function (which projects multi-objective values to scalar ones), and the type of policies considered. We show how these factors determine the nature of an optimal solution, which can be a single policy, a convex hull, or a Pareto front. Using this taxonomy, we survey the literature on multi-objective methods for planning and learning. Finally, we discuss key applications of such methods and outline opportunities for future work.