Directly Aligning the Full Diffusion Trajectory with Fine-Grained Human Preference

Recent studies have demonstrated the effectiveness of directly aligning diffusion models with human preferences using differentiable reward. However, they exhibit two primary challenges: (1) they rely on multistep denoising with gradient computation for reward scoring, which is computationally expensive, thus restricting optimization to only a few diffusion steps; (2) they often need continuous offline adaptation of reward models in order to achieve desired aesthetic quality, such as photorealism or precise lighting effects. To address the limitation of multistep denoising, we propose Direct-Align, a method that predefines a noise prior to effectively recover original images from any time steps via interpolation, leveraging the equation that diffusion states are interpolations between noise and target images, which effectively avoids over-optimization in late timesteps. Furthermore, we introduce Semantic Relative Preference Optimization (SRPO), in which rewards are formulated as text-conditioned signals. This approach enables online adjustment of rewards in response to positive and negative prompt augmentation, thereby reducing the reliance on offline reward fine-tuning. By fine-tuning the FLUX model with optimized denoising and online reward adjustment, we improve its human-evaluated realism and aesthetic quality by over 3x.

Hunyuan-Game: Industrial-grade Intelligent Game Creation Model

Intelligent game creation represents a transformative advancement in game development, utilizing generative artificial intelligence to dynamically generate and enhance game content. Despite notable progress in generative models, the comprehensive synthesis of high-quality game assets, including both images and videos, remains a challenging frontier. To create high-fidelity game content that simultaneously aligns with player preferences and significantly boosts designer efficiency, we present Hunyuan-Game, an innovative project designed to revolutionize intelligent game production. Hunyuan-Game encompasses two primary branches: image generation and video generation. The image generation component is built upon a vast dataset comprising billions of game images, leading to the development of a group of customized image generation models tailored for game scenarios: (1) General Text-to-Image Generation. (2) Game Visual Effects Generation, involving text-to-effect and reference image-based game visual effect generation. (3) Transparent Image Generation for characters, scenes, and game visual effects. (4) Game Character Generation based on sketches, black-and-white images, and white models. The video generation component is built upon a comprehensive dataset of millions of game and anime videos, leading to the development of five core algorithmic models, each targeting critical pain points in game development and having robust adaptation to diverse game video scenarios: (1) Image-to-Video Generation. (2) 360 A/T Pose Avatar Video Synthesis. (3) Dynamic Illustration Generation. (4) Generative Video Super-Resolution. (5) Interactive Game Video Generation. These image and video generation models not only exhibit high-level aesthetic expression but also deeply integrate domain-specific knowledge, establishing a systematic understanding of diverse game and anime art styles.

Augmenting Online RL with Offline Data is All You Need: A Unified Hybrid RL Algorithm Design and Analysis

This paper investigates a hybrid learning framework for reinforcement learning (RL) in which the agent can leverage both an offline dataset and online interactions to learn the optimal policy. We present a unified algorithm and analysis and show that augmenting confidence-based online RL algorithms with the offline dataset outperforms any pure online or offline algorithm alone and achieves state-of-the-art results under two learning metrics, i.e., sub-optimality gap and online learning regret. Specifically, we show that our algorithm achieves a sub-optimality gap $\tilde{O}(\sqrt{1/(N_0/\mathtt{C}(\pi^*|\rho)+N_1}) )$, where $\mathtt{C}(\pi^*|\rho)$ is a new concentrability coefficient, $N_0$ and $N_1$ are the numbers of offline and online samples, respectively. For regret minimization, we show that it achieves a constant $\tilde{O}( \sqrt{N_1/(N_0/\mathtt{C}(\pi^{-}|\rho)+N_1)} )$ speed-up compared to pure online learning, where $\mathtt{C}(\pi^-|\rho)$ is the concentrability coefficient over all sub-optimal policies. Our results also reveal an interesting separation on the desired coverage properties of the offline dataset for sub-optimality gap minimization and regret minimization. We further validate our theoretical findings in several experiments in special RL models such as linear contextual bandits and Markov decision processes (MDPs).

A Shared Low-Rank Adaptation Approach to Personalized RLHF

Reinforcement Learning from Human Feedback (RLHF) has emerged as a pivotal technique for aligning artificial intelligence systems with human values, achieving remarkable success in fine-tuning large language models. However, existing RLHF frameworks often assume that human preferences are relatively homogeneous and can be captured by a single, unified reward model. This assumption overlooks the inherent diversity and heterogeneity across individuals, limiting the adaptability of RLHF to personalized scenarios and risking misalignments that can diminish user satisfaction and trust in AI systems. In this paper, we address these challenges by introducing Low-Rank Adaptation (LoRA) into the personalized RLHF framework. We apply LoRA in the the aggregated parameter space of all personalized reward functions, thereby enabling efficient learning of personalized reward models from potentially limited local datasets. Our approach exploits potential shared structures among the local ground-truth reward models while allowing for individual adaptation, without relying on restrictive assumptions about shared representations as in prior works. We further establish sample complexity guarantees for our method. Theoretical analysis demonstrates the effectiveness of the proposed approach in capturing both shared and individual-specific structures within heterogeneous human preferences, addressing the dual challenge of personalization requirements and practical data constraints. Experimental results on real-world datasets corroborate the efficiency of our algorithm in the personalized RLHF setting.

Evaluating Membership Inference Attacks and Defenses in Federated Learning

Membership Inference Attacks (MIAs) pose a growing threat to privacy preservation in federated learning. The semi-honest attacker, e.g., the server, may determine whether a particular sample belongs to a target client according to the observed model information. This paper conducts an evaluation of existing MIAs and corresponding defense strategies. Our evaluation on MIAs reveals two important findings about the trend of MIAs. Firstly, combining model information from multiple communication rounds (Multi-temporal) enhances the overall effectiveness of MIAs compared to utilizing model information from a single epoch. Secondly, incorporating models from non-target clients (Multi-spatial) significantly improves the effectiveness of MIAs, particularly when the clients' data is homogeneous. This highlights the importance of considering the temporal and spatial model information in MIAs. Next, we assess the effectiveness via privacy-utility tradeoff for two type defense mechanisms against MIAs: Gradient Perturbation and Data Replacement. Our results demonstrate that Data Replacement mechanisms achieve a more optimal balance between preserving privacy and maintaining model utility. Therefore, we recommend the adoption of Data Replacement methods as a defense strategy against MIAs. Our code is available in https://github.com/Liar-Mask/FedMIA.

P-Bench: A Multi-level Privacy Evaluation Benchmark for Language Models

The rapid development of language models (LMs) brings unprecedented accessibility and usage for both models and users. On the one hand, powerful LMs, trained with massive textual data, achieve state-of-the-art performance over numerous downstream NLP tasks. On the other hand, more and more attention is paid to unrestricted model accesses that may bring malicious privacy risks of data leakage. To address these issues, many recent works propose privacy-preserving language models (PPLMs) with differential privacy (DP). Unfortunately, different DP implementations make it challenging for a fair comparison among existing PPLMs. In this paper, we present P-Bench, a multi-perspective privacy evaluation benchmark to empirically and intuitively quantify the privacy leakage of LMs. Instead of only protecting and measuring the privacy of protected data with DP parameters, P-Bench sheds light on the neglected inference data privacy during actual usage. P-Bench first clearly defines multi-faceted privacy objectives during private fine-tuning. Then, P-Bench constructs a unified pipeline to perform private fine-tuning. Lastly, P-Bench performs existing privacy attacks on LMs with pre-defined privacy objectives as the empirical evaluation results. The empirical attack results are used to fairly and intuitively evaluate the privacy leakage of various PPLMs. We conduct extensive experiments on three datasets of GLUE for mainstream LMs.

ChatCounselor: A Large Language Models for Mental Health Support

This paper presents ChatCounselor, a large language model (LLM) solution designed to provide mental health support. Unlike generic chatbots, ChatCounselor is distinguished by its foundation in real conversations between consulting clients and professional psychologists, enabling it to possess specialized knowledge and counseling skills in the field of psychology. The training dataset, Psych8k, was constructed from 260 in-depth interviews, each spanning an hour. To assess the quality of counseling responses, the counseling Bench was devised. Leveraging GPT-4 and meticulously crafted prompts based on seven metrics of psychological counseling assessment, the model underwent evaluation using a set of real-world counseling questions. Impressively, ChatCounselor surpasses existing open-source models in the counseling Bench and approaches the performance level of ChatGPT, showcasing the remarkable enhancement in model capability attained through high-quality domain-specific data.

Robot motor learning shows emergence of frequency-modulated, robust swimming with an invariant Strouhal-number

Fish locomotion emerges from a diversity of interactions among deformable structures, surrounding fluids and neuromuscular activations, i.e., fluid-structure interactions (FSI) controlled by fish's motor systems. Previous studies suggested that such motor-controlled FSI may possess embodied traits. However, their implications in motor learning, neuromuscular control, gait generation, and swimming performance remain to be uncovered. Using robot models, we studied how swimming behaviours emerged from the FSI and the embodied traits. We developed modular robots with various designs and used Central Pattern Generators (CPGs) to control the torque acting on robot body. We used reinforcement learning to learn CPG parameters to maximize the swimming speed. The results showed that motor frequency converged faster than other parameters, and the emergent swimming gaits were robust against disruptions applied to motor control. For all robots and frequencies tested, swimming speed was proportional to the mean undulation velocity of body and caudal-fin combined, yielding an invariant, undulation-based Strouhal number. The Strouhal number also revealed two fundamental classes of undulatory swimming in both biological and robotic fishes. The robot actuators also demonstrated diverse functions as motors, virtual springs, and virtual masses. These results provide novel insights into the embodied traits of motor-controlled FSI for fish-inspired locomotion.

Near-optimal Conservative Exploration in Reinforcement Learning under Episode-wise Constraints

This paper investigates conservative exploration in reinforcement learning where the performance of the learning agent is guaranteed to be above a certain threshold throughout the learning process. It focuses on the tabular episodic Markov Decision Process (MDP) setting that has finite states and actions. With the knowledge of an existing safe baseline policy, an algorithm termed as StepMix is proposed to balance the exploitation and exploration while ensuring that the conservative constraint is never violated in each episode with high probability. StepMix features a unique design of a mixture policy that adaptively and smoothly interpolates between the baseline policy and the optimistic policy. Theoretical analysis shows that StepMix achieves near-optimal regret order as in the constraint-free setting, indicating that obeying the stringent episode-wise conservative constraint does not compromise the learning performance. Besides, a randomization-based EpsMix algorithm is also proposed and shown to achieve the same performance as StepMix. The algorithm design and theoretical analysis are further extended to the setting where the baseline policy is not given a priori but must be learned from an offline dataset, and it is proved that similar conservative guarantee and regret can be achieved if the offline dataset is sufficiently large. Experiment results corroborate the theoretical analysis and demonstrate the effectiveness of the proposed conservative exploration strategies.

Random Smoothing Regularization in Kernel Gradient Descent Learning

Random smoothing data augmentation is a unique form of regularization that can prevent overfitting by introducing noise to the input data, encouraging the model to learn more generalized features. Despite its success in various applications, there has been a lack of systematic study on the regularization ability of random smoothing. In this paper, we aim to bridge this gap by presenting a framework for random smoothing regularization that can adaptively and effectively learn a wide range of ground truth functions belonging to the classical Sobolev spaces. Specifically, we investigate two underlying function spaces: the Sobolev space of low intrinsic dimension, which includes the Sobolev space in $D$-dimensional Euclidean space or low-dimensional sub-manifolds as special cases, and the mixed smooth Sobolev space with a tensor structure. By using random smoothing regularization as novel convolution-based smoothing kernels, we can attain optimal convergence rates in these cases using a kernel gradient descent algorithm, either with early stopping or weight decay. It is noteworthy that our estimator can adapt to the structural assumptions of the underlying data and avoid the curse of dimensionality. This is achieved through various choices of injected noise distributions such as Gaussian, Laplace, or general polynomial noises, allowing for broad adaptation to the aforementioned structural assumptions of the underlying data. The convergence rate depends only on the effective dimension, which may be significantly smaller than the actual data dimension. We conduct numerical experiments on simulated data to validate our theoretical results.