亚洲男人的天堂2018av,欧美草比,久久久久久免费视频精选,国色天香在线看免费,久久久久亚洲av成人片仓井空

Traditional reinforcement learning from human feedback (RLHF) approaches relying on parametric models like the Bradley-Terry model fall short in capturing the intransitivity and irrationality in human preferences. Recent advancements suggest that directly working with preference probabilities can yield a more accurate reflection of human preferences, enabling more flexible and accurate language model alignment. In this paper, we propose a self-play-based method for language model alignment, which treats the problem as a constant-sum two-player game aimed at identifying the Nash equilibrium policy. Our approach, dubbed \textit{Self-Play Preference Optimization} (SPPO), approximates the Nash equilibrium through iterative policy updates and enjoys theoretical convergence guarantee. Our method can effectively increase the log-likelihood of the chosen response and decrease that of the rejected response, which cannot be trivially achieved by symmetric pairwise loss such as Direct Preference Optimization (DPO) and Identity Preference Optimization (IPO). In our experiments, using only 60k prompts (without responses) from the UltraFeedback dataset and without any prompt augmentation, by leveraging a pre-trained preference model PairRM with only 0.4B parameters, SPPO can obtain a model from fine-tuning Mistral-7B-Instruct-v0.2 that achieves the state-of-the-art length-controlled win-rate of 28.53% against GPT-4-Turbo on AlpacaEval 2.0. It also outperforms the (iterative) DPO and IPO on MT-Bench and the Open LLM Leaderboard. Notably, the strong performance of SPPO is achieved without additional external supervision (e.g., responses, preferences, etc.) from GPT-4 or other stronger language models.

The alignments of reasoning abilities between smaller and larger Language Models are largely conducted via Supervised Fine-Tuning (SFT) using demonstrations generated from robust Large Language Models (LLMs). Although these approaches deliver more performant models, they do not show sufficiently strong generalization ability as the training only relies on the provided demonstrations. In this paper, we propose the Self-refine Instruction-tuning method that elicits Smaller Language Models to self-refine their abilities. Our approach is based on a two-stage process, where reasoning abilities are first transferred between LLMs and Small Language Models (SLMs) via Instruction-tuning on demonstrations provided by LLMs, and then the instructed models Self-refine their abilities through preference optimization strategies. In particular, the second phase operates refinement heuristics based on the Direct Preference Optimization algorithm, where the SLMs are elicited to deliver a series of reasoning paths by automatically sampling the generated responses and providing rewards using ground truths from the LLMs. Results obtained on commonsense and math reasoning tasks show that this approach significantly outperforms Instruction-tuning in both in-domain and out-domain scenarios, aligning the reasoning abilities of Smaller and Larger Language Models.

We introduce AdaMoLE, a novel method for fine-tuning large language models (LLMs) through an Adaptive Mixture of Low-Rank Adaptation (LoRA) Experts. Moving beyond conventional methods that employ a static top-k strategy for activating experts, AdaMoLE dynamically adjusts the activation threshold using a dedicated threshold network, adaptively responding to the varying complexities of different tasks. By replacing a single LoRA in a layer with multiple LoRA experts and integrating a gating function with the threshold mechanism, AdaMoLE effectively selects and activates the most appropriate experts based on the input context. Our extensive evaluations across a variety of commonsense reasoning and natural language processing tasks show that AdaMoLE exceeds baseline performance. This enhancement highlights the advantages of AdaMoLE's adaptive selection of LoRA experts, improving model effectiveness without a corresponding increase in the expert count. The experimental validation not only confirms AdaMoLE as a robust approach for enhancing LLMs but also suggests valuable directions for future research in adaptive expert selection mechanisms, potentially broadening the scope for optimizing model performance across diverse language processing tasks.

Fairness is steadily becoming a crucial requirement of Machine Learning (ML) systems. A particularly important notion is subgroup fairness, i.e., fairness in subgroups of individuals that are defined by more than one attributes. Identifying bias in subgroups can become both computationally challenging, as well as problematic with respect to comprehensibility and intuitiveness of the finding to end users. In this work we focus on the latter aspects; we propose an explainability method tailored to identifying potential bias in subgroups and visualizing the findings in a user friendly manner to end users. In particular, we extend the ALE plots explainability method, proposing FALE (Fairness aware Accumulated Local Effects) plots, a method for measuring the change in fairness for an affected population corresponding to different values of a feature (attribute). We envision FALE to function as an efficient, user friendly, comprehensible and reliable first-stage tool for identifying subgroups with potential bias issues.

Efficient and robust anisotropic mesh adaptation is crucial for Computational Fluid Dynamics (CFD) simulations. The CFD Vision 2030 Study highlights the pressing need for this technology, particularly for simulations targeting supercomputers. This work applies a fine-grained speculative approach to anisotropic mesh operations. Our implementation exhibits more than 90% parallel efficiency on a multi-core node. Additionally, we evaluate our method within an adaptive pipeline for a spectrum of publicly available test-cases that includes both analytically derived and error-based fields. For all test-cases, our results are in accordance with published results in the literature. Support for CAD-based data is introduced, and its effectiveness is demonstrated on one of NASA's High-Lift prediction workshop cases.

We propose a time-dependent Advection Reaction Diffusion (ARD) $N$-species competition model to investigate the Stocking and Harvesting (SH) effect on population dynamics. For ongoing analysis, we explore the outcomes of a competition between two competing species in a heterogeneous environment under no-flux boundary conditions, meaning no individual can cross the boundaries. We establish results concerning the existence, uniqueness, and positivity of the solution. As a continuation, we propose, analyze, and test two novel fully discrete decoupled linearized algorithms for a nonlinearly coupled ARD $N$-species competition model with SH effort. The time-stepping algorithms are first and second order accurate in time and optimally accurate in space. Stability and optimal convergence theorems of the decoupled schemes are proved rigorously. We verify the predicted convergence rates of our analysis and the efficacy of the algorithms using numerical experiments and synthetic data for analytical test problems. We also study the effect of harvesting or stocking and diffusion parameters on the evolution of species population density numerically and observe the coexistence scenario subject to optimal stocking or harvesting.

Meeting the strict Quality of Service (QoS) requirements of terminals has imposed a signiffcant challenge on Multiaccess Edge Computing (MEC) systems, due to the limited multidimensional resources. To address this challenge, we propose a collaborative MEC framework that facilitates resource sharing between the edge servers, and with the aim to maximize the long-term QoS and reduce the cache switching cost through joint optimization of service caching, collaborative offfoading, and computation and communication resource allocation. The dual timescale feature and temporal recurrence relationship between service caching and other resource allocation make solving the problem even more challenging. To solve it, we propose a deep reinforcement learning (DRL)-based dual timescale scheme, called DGL-DDPG, which is composed of a short-term genetic algorithm (GA) and a long short-term memory network-based deep deterministic policy gradient (LSTM-DDPG). In doing so, we reformulate the optimization problem as a Markov decision process (MDP) where the small-timescale resource allocation decisions generated by an improved GA are taken as the states and input into a centralized LSTM-DDPG agent to generate the service caching decision for the large-timescale. Simulation results demonstrate that our proposed algorithm outperforms the baseline algorithms in terms of the average QoS and cache switching cost.

We introduce ObjectAdd, a training-free diffusion modification method to add user-expected objects into user-specified area. The motive of ObjectAdd stems from: first, describing everything in one prompt can be difficult, and second, users often need to add objects into the generated image. To accommodate with real world, our ObjectAdd maintains accurate image consistency after adding objects with technical innovations in: (1) embedding-level concatenation to ensure correct text embedding coalesce; (2) object-driven layout control with latent and attention injection to ensure objects accessing user-specified area; (3) prompted image inpainting in an attention refocusing & object expansion fashion to ensure rest of the image stays the same. With a text-prompted image, our ObjectAdd allows users to specify a box and an object, and achieves: (1) adding object inside the box area; (2) exact content outside the box area; (3) flawless fusion between the two areas

Generalization to out-of-distribution (OOD) data is a capability natural to humans yet challenging for machines to reproduce. This is because most learning algorithms strongly rely on the i.i.d.~assumption on source/target data, which is often violated in practice due to domain shift. Domain generalization (DG) aims to achieve OOD generalization by using only source data for model learning. Since first introduced in 2011, research in DG has made great progresses. In particular, intensive research in this topic has led to a broad spectrum of methodologies, e.g., those based on domain alignment, meta-learning, data augmentation, or ensemble learning, just to name a few; and has covered various vision applications such as object recognition, segmentation, action recognition, and person re-identification. In this paper, for the first time a comprehensive literature review is provided to summarize the developments in DG for computer vision over the past decade. Specifically, we first cover the background by formally defining DG and relating it to other research fields like domain adaptation and transfer learning. Second, we conduct a thorough review into existing methods and present a categorization based on their methodologies and motivations. Finally, we conclude this survey with insights and discussions on future research directions.

Multiple instance learning (MIL) is a powerful tool to solve the weakly supervised classification in whole slide image (WSI) based pathology diagnosis. However, the current MIL methods are usually based on independent and identical distribution hypothesis, thus neglect the correlation among different instances. To address this problem, we proposed a new framework, called correlated MIL, and provided a proof for convergence. Based on this framework, we devised a Transformer based MIL (TransMIL), which explored both morphological and spatial information. The proposed TransMIL can effectively deal with unbalanced/balanced and binary/multiple classification with great visualization and interpretability. We conducted various experiments for three different computational pathology problems and achieved better performance and faster convergence compared with state-of-the-art methods. The test AUC for the binary tumor classification can be up to 93.09% over CAMELYON16 dataset. And the AUC over the cancer subtypes classification can be up to 96.03% and 98.82% over TCGA-NSCLC dataset and TCGA-RCC dataset, respectively.