dvd
Amazon drops early Black Friday deals on Pokémon DVDs: Shop classic seasons starting under 15.99
Look Up Mashable's Best: E-readers, robovacs, laptops, earbuds, smart home and more Mashable Selects Say More Trending Now Gift Ideas For Everyone On Your List Safety Net Mashable @Home In My Bag Versus Creator Playbook Getting Technical with Mashable All Series Adult you has an Amazon account and zero impulse control. Soumya is a deals writer who covers consumer tech, shopping deals, and the products people use every day. With experience writing about everything from AI tools and software to smartphones and home gadgets, she enjoys breaking down product research into clear, useful recommendations. When she's not tracking deals, she's usually comparing products, digging through reviews, and figuring out what actually makes a purchase worth it. All products featured here are independently selected by our editors and writers.
Heated Rivalry box set deal: Preorders for the wild Season 1 Collectors Edition box set are on sale at Amazon
Look Up Mashable's Best: E-readers, robovacs, laptops, earbuds, smart home and more Say More Safety Net Creator Hub Versus Gift Ideas For Everyone On Your List Mashable Selects Switch Off Trending Now In My Bag VidCon with Mashable All Series'Heated Rivalry' box set deal: Preorders for the wild Season 1 Collector's Edition box set are on sale at Amazon Lauren Allain is a freelance journalist covering deals at Mashable. She graduated from Western Washington University with a B.A. in journalism and holds an M.B.A from Webster Leiden. You can find more of her work online from publications including Reader's Digest, U.S. News & World Report, Seattle Refined, and more. When she's not writing, Lauren prefers to be outside hiking, bouldering, swimming, or searching for the perfect location for all three. All products featured here are independently selected by our editors and writers.
Vicinity-Guided Discriminative Latent Diffusion for Privacy-Preserving Domain Adaptation
Recent work on latent diffusion models (LDMs) has focused almost exclusively on generative tasks, leaving their potential for discriminative transfer largely unexplored. We introduce Discriminative Vicinity Diffusion (DVD), a novel LDM-based framework for a more practical variant of source-free domain adaptation (SFDA): the source provider may share not only a pre-trained classifier but also an auxiliary latent diffusion module, trained once on the source data and never exposing raw source samples. DVD encodes each source feature's label information into its latent vicinity by fitting a Gaussian prior over its k-nearest neighbors and training the diffusion network to "drift" noisy samples back to label-consistent representations. During adaptation, we sample from each target feature's latent vicinity, apply the frozen diffusion module to generate source-like cues, and use a simple InfoNCE loss to align the target encoder to these cues, explicitly transferring decision boundaries without source access. Across standard SFDA benchmarks, DVD outperforms state-of-the-art methods. We further show that the same latent diffusion module enhances the source classifier's accuracy on in-domain data and boosts performance in supervised classification and domain generalization experiments. DVD thus reinterprets LDMs as practical, privacy-preserving bridges for explicit knowledge transfer, addressing a core challenge in source-free domain adaptation that prior methods have yet to solve.
Vicinity-Guided Discriminative Latent Diffusion for Privacy-Preserving Domain Adaptation
Wang, Jing, Bae, Wonho, Chen, Jiahong, Wang, Wenxu, Noh, Junhyug
Recent work on latent diffusion models (LDMs) has focused almost exclusively on generative tasks, leaving their potential for discriminative transfer largely unexplored. We introduce Discriminative Vicinity Diffusion (DVD), a novel LDM-based framework for a more practical variant of source-free domain adaptation (SFDA): the source provider may share not only a pre-trained classifier but also an auxiliary latent diffusion module, trained once on the source data and never exposing raw source samples. DVD encodes each source feature's label information into its latent vicinity by fitting a Gaussian prior over its k-nearest neighbors and training the diffusion network to drift noisy samples back to label-consistent representations. During adaptation, we sample from each target feature's latent vicinity, apply the frozen diffusion module to generate source-like cues, and use a simple InfoNCE loss to align the target encoder to these cues, explicitly transferring decision boundaries without source access. Across standard SFDA benchmarks, DVD outperforms state-of-the-art methods. We further show that the same latent diffusion module enhances the source classifier's accuracy on in-domain data and boosts performance in supervised classification and domain generalization experiments. DVD thus reinterprets LDMs as practical, privacy-preserving bridges for explicit knowledge transfer, addressing a core challenge in source-free domain adaptation that prior methods have yet to solve.
Review for NeurIPS paper: Effective Diversity in Population Based Reinforcement Learning
Weaknesses: The paper may need to be improved to address a few important issues, as detailed below. Why is it important to enhance population-wide behavioral diversity? Intuitively I can understand the potential benefits related to deep exploration and learning stability. However, theoretically I cannot link the benefits straightforwardly to the proposed use of kernel function and the kernel matrix determinant. Theorem 3.3 states that when lambda is set properly, the population will contain M distinct optimal policies.
Markov Chain of Thought for Efficient Mathematical Reasoning
Yang, Wen, Fan, Kai, Liao, Minpeng
Chain of Thought (CoT) of multi-step benefits from the logical structure of the reasoning steps and task-specific actions, significantly enhancing the mathematical reasoning capabilities of large language models. As the prevalence of long CoT, the number of reasoning steps exceeds manageable token limits and leads to higher computational demands. Inspired by the fundamental logic of human cognition, ``derive, then reduce'', we conceptualize the standard multi-step CoT as a novel Markov Chain of Thought (MCoT). In this study, we consider the mathematical reasoning task, defining each reasoning step as text accompanied by a Python code snippet. To facilitate a longer reasoning path, self-correction is enabled through interactions with the code interpreter. Our MCoT aims to compress previous reasoning steps into a simplified question, enabling efficient next-step inference without relying on a lengthy KV cache. In our experiments, we curate the \texttt{MCoTInstruct} dataset, and the empirical results indicate that MCoT not only significantly enhances efficiency but also maintains comparable accuracy. While much remains to be explored, this work paves the way for exploring the long CoT reasoning abilities of LLMs.
PersonaMath: Enhancing Math Reasoning through Persona-Driven Data Augmentation
Luo, Jing, Luo, Run, Chen, Longze, Zhu, Liang, Ao, Chang, Li, Jiaming, Chen, Yukun, Cheng, Xin, Yang, Wen, Su, Jiayuan, Li, Chengming, Yang, Min
While closed-source Large Language Models (LLMs) demonstrate strong mathematical problem-solving abilities, open-source models continue to struggle with such tasks. To bridge this gap, we propose a data augmentation approach and introduce PersonaMathQA, a dataset derived from MATH and GSM8K, on which we train the PersonaMath models. Our approach consists of two stages: the first stage is learning from Persona Diversification, and the second stage is learning from Reflection. In the first stage, we regenerate detailed chain-of-thought (CoT) solutions as instructions using a closed-source LLM and introduce a novel personadriven data augmentation technique to enhance the dataset's quantity and diversity. In the second stage, we incorporate reflection to fully leverage more challenging and valuable questions. Evaluation of our PersonaMath models on MATH and GSM8K reveals that the PersonaMath-7B model (based on LLaMA-2-7B) achieves an accuracy of 24.2% on MATH and 68.7% on GSM8K, surpassing all baseline methods and achieving state-of-the-art performance. Notably, our dataset contains only 70.3K data points--merely 17.8% of MetaMathQA and 27% of MathInstruct--yet our model outperforms these baselines, demonstrating the high quality and diversity of our dataset, which enables more efficient model training. "There are a thousand Hamlets in a thousand people's eyes" Among these tasks, solving math problems stands out as particularly challenging due to its complexity and the requirement for multi-step reasoning to reach a solution. While some closed-source models, such as GPT-4o (OpenAI, 2024a), Claude 3.5 Sonnet (Anthropic, 2024), and Gemini 1.5 Pro (Reid et al., 2024), have demonstrated strong math-solving capabilities, current open-source models (e.g., LLaMA (Touvron et al., 2023; Dubey et al., 2024)) continue to struggle in this area. Therefore, enhancing the math problem-solving abilities of open-source models is a prominent desiderata. A widely adopted and effective approach for improving the math-solving capabilities of open-source models is fine-tuning, owing to the accessibility of their weights (Yuan et al., 2023; Yue et al., 2023; The method consists of two stages: Stage 1 (top) and Stage 2 (bottom). Stage 1 focuses on using closed-source LLMs to automatically generate detailed CoT solutions and apply our persona-driven rewriting method to rephrase the questions.
VarBench: Robust Language Model Benchmarking Through Dynamic Variable Perturbation
Qian, Kun, Wan, Shunji, Tang, Claudia, Wang, Youzhi, Zhang, Xuanming, Chen, Maximillian, Yu, Zhou
As large language models achieve impressive scores on traditional benchmarks, an increasing number of researchers are becoming concerned about benchmark data leakage during pre-training, commonly known as the data contamination problem. To ensure fair evaluation, recent benchmarks release only the training and validation sets, keeping the test set labels closed-source. They require anyone wishing to evaluate his language model to submit the model's predictions for centralized processing and then publish the model's result on their leaderboard. However, this submission process is inefficient and prevents effective error analysis. To address this issue, we propose to variabilize benchmarks and evaluate language models dynamically. Specifically, we extract variables from each test case and define a value range for each variable. For each evaluation, we sample new values from these value ranges to create unique test cases, thus ensuring a fresh evaluation each time. We applied this variable perturbation method to four datasets: GSM8K, ARC, CommonsenseQA, and TruthfulQA, which cover mathematical generation and multiple-choice tasks. Our experimental results demonstrate that this approach provides a more accurate assessment of the true capabilities of language models, effectively mitigating the contamination problem.
How disruption can lead to innovation in how you conduct your business
When Marc Randolph and Reed Hastings launched the world's first online DVD-rental service in California in 1997, it was far from the behemoth that Netflix is today. Customers had fewer than 1,000 titles to choose from, and they had to pay separately for every DVD they rented. The founders nearly sold the company to Amazon for around US$15 million, and when they tried to sell it to Blockbuster forUS$50 million a few years later, Blockbuster's CEO was reported to have said to Netflix, "The dotcom hysteria is completely overblown." Fast forward two decades, and the company now has more than 200 million subscribers around the world and more than US$20 billion in annual revenue. The simple answer is'reinvention' – a requirement that can present itself to any business, but which not all business leaders have the courage or forethought to embrace.
A Brief Introduction to Recommendation Systems
Have you ever wondered how apps like Netflix or Spotify decide which movie or songs you're likely to prefer watching or listening to? Seems like magic, doesn't it? For instance, a lot of data is being mined and multiple complicated algorithms are developed by data science professionals in an attempt to make predictions more accurate. It is not magic but "machine learning." Machine learning is what allows the system to determine the movies and songs most relevant to your liking.