Don't miss this: Double your impact! We're able to stand strong because we're funded by readers like you. Support journalism that doesn't flinch. Don't miss this: Tomorrow is the final day of our $50,000 match We're able to stand strong because we're funded by readers like you. Support journalism that doesn't flinch.

As AI becomes more personal, e.g., Agentic AI, there is an increasing need for personalizing models for various use cases. Personalized federated learning (PFL) enables each client to collaboratively leverage other clients' knowledge for better adaptation to the task of interest, without privacy risks. Despite its potential, existing PFL methods remain confined to rather simplified scenarios where data and models are the same across clients. To move towards realistic scenarios, we propose FedMosaic, a method that jointly addresses data and model heterogeneity with a task-relevance-aware model aggregation strategy to reduce parameter interference, and a dimension-invariant module that enables knowledge sharing across heterogeneous architectures without huge computational cost. To mimic the real-world task diversity, we propose a multi-modal PFL benchmark spanning 40 distinct tasks with distribution shifts over time. The empirical study shows that FedMosaic outperforms the state-of-the-art PFL methods, excelling in both personalization and generalization capabilities under challenging, realistic scenarios.

Ab s tract -- The proliferation of Text - to - Music (TTM) platforms has democratized music creation, letting users effortlessly generat e high - quality compositions . However, this innovation has also introduced challenges to musicians and the music in dustry . T his research focuses on utilizing the FakeMusicCaps dataset to address the challenge of detecting AI - generated songs by classifying the audio as deepfake or human. To simulate a real - world adversarial entity tempo stretching and pitch shifting modifications were applied to the dataset . Mel Spectrograms were generated from the resulting datasets, w hich were then used to train and test a convolutional neural network. This paper also explores the ethical and societal implications of TTM platforms, suggesting that detection systems developed and employed with care are a necessary tool to safeguard musicians and foster the positive potential of TTM plat forms and gen erative AI in music . Rapid a dvances in g e nerative AI have caused the creat ive landscape to be u pended, enabling almost anyone to easily create music that can be hard to distinguish from human - ma de compositions . AI - generated music is part of a wider classification of AI - generated media and art that falls unde r the category of " deepfake " .

Recent studies have demonstrated the strong empirical performance of diffusion models on discrete sequences (i.e., discrete diffusion models) across domains from natural language to biological sequence generation. For example, in the protein inverse folding task, where the goal is to generate a protein sequence from a given backbone structure, conditional diffusion models have achieved impressive results in generating natural-like sequences that fold back into the original structure. However, practical design tasks often require not only modeling a conditional distribution but also optimizing specific task objectives. For instance, in the inverse folding task, we may prefer protein sequences with high stability. To address this, we consider the scenario where we have pre-trained discrete diffusion models that can generate natural-like sequences, as well as reward models that map sequences to task objectives. We then formulate the reward maximization problem within discrete diffusion models, analogous to reinforcement learning (RL), while minimizing the KL divergence against pretrained diffusion models to preserve naturalness. To solve this RL problem, we propose a novel algorithm, DRAKES, that enables direct backpropagation of rewards through entire trajectories generated by diffusion models, by making the originally nondifferentiable trajectories differentiable using the Gumbel-Softmax trick. Our theoretical analysis indicates that our approach can generate sequences that are both natural-like (i.e., have a high probability under a pretrained model) and yield high rewards. While similar tasks have been recently explored in diffusion models for continuous domains, our work addresses unique algorithmic and theoretical challenges specific to discrete diffusion models, which arise from their foundation in continuous-time Markov chains rather than Brownian motion. Finally, we demonstrate the effectiveness of our algorithm in generating DNA and protein sequences that optimize enhancer activity and protein stability, respectively, important tasks for gene therapies and protein-based therapeutics. Diffusion models have gained widespread recognition as effective generative models in continuous spaces, such as image and video generation (Song et al., 2020; Ho et al., 2022). Inspired by seminal works (e.g., Austin et al. (2021); Campbell et al. (2022); Sun et al. (2022)), recent studies (Lou et al., 2023; Shi et al., 2024; Sahoo et al., 2024) have shown that diffusion models are also highly effective in discrete spaces, including natural language and biological sequence generation (DNA, RNA, proteins). Work mainly done during an internship at Genentech.

Physics-based simulation is essential for developing and evaluating robot manipulation policies, particularly in scenarios involving deformable objects and complex contact interactions. However, existing simulators often struggle to balance computational efficiency with numerical accuracy, especially when modeling deformable materials with frictional contact constraints. We introduce an efficient subspace representation for the Incremental Potential Contact (IPC) method, leveraging model reduction to decrease the number of degrees of freedom. Our approach decouples simulation complexity from the resolution of the input model by representing elasticity in a low-resolution subspace while maintaining collision constraints on an embedded high-resolution surface. Our barrier formulation ensures intersection-free trajectories and configurations regardless of material stiffness, time step size, or contact severity. We validate our simulator through quantitative experiments with a soft bubble gripper grasping and qualitative demonstrations of placing a plate on a dish rack. The results demonstrate our simulator's efficiency, physical accuracy, computational stability, and robust handling of frictional contact, making it well-suited for generating demonstration data and evaluating downstream robot training applications.

Four days after the attempt on his life, the voice of Donald Trump booms from the speakers in Montego Bay, Jamaica: "If they needed an assassin, they should have sent for Bodyguard … about to commit a quadruple murder at Sumfest in Montego Bay." The audience are taken by surprise, having been primed for a reggae riddim to drop, and laugh. The Bodyguard crew have just taken to the stage at Sumfest Global Sound Clash, a musical gladiatorial contest where sound systems battle against one another with creative mixing, hyped-up MCs and exclusive – often incendiary – recordings featuring star guests and in-jokes. AI vocalists such as this fake Trump, however, are sending shockwaves through a decades-old musical tradition in which authenticity and originality are paramount, and sound systems pay premium rates to artists to get vocals for clashes. "AI is going to mash up the industry," says Fabian Anderson, a dub agent who liaises between artists and sound systems to secure those exclusive tracks.

As if he didn't have enough to deal with amid his beef with Kendrick Lamar (or perhaps to distract from it), Drake showed up on a remix of parody rapper Snowd4y's cover of Plain White T's "Hey There Delilah," called "Wah Gwan Delilah," that has everyone … perplexed? Let's walk through this together, it's a mess. It had what appeared to be Drake joining the comedian in a series of quips about women and name-checks of Toronto landmarks like the Yonge-Dundas Square mall. As the track spread, it made its way to the Plain White T's themselves, who posted a video on X and TikTok with the caption "too stunned to speak." Frontman Tom Higgenson also says "it's crazy that everybody thinks that it's real," seemingly referencing early rumors that Drake's lyrics were generated using artificial intelligence.

Drake and Kendrick Lamar's ongoing beef has left their devoted fans utterly divided over who's winning. The rappers have released several diss songs against each other and their fan bases are certain their team is winning. To try and strip biases out of the debate, we asked artificial intelligence chatbots who is winning the ongoing feud - and it produced some surprising insights. Three out of four AI chatbots remained politically correct when addressing which rapper is winning the beef, calling it'subjective' and saying it is up to the fans to decide. But Meta's AI bot said Kendrick had a slight edge in the beef so far. Gemini called Drake a'commercial powerhouse with numerous hit singles and albums that have topped the charts,' but said who is winning the feud remains subjective'Ultimately, the winner of the Drake versus Kendrick Lamar beef is subjective and depends on personal preference,' said DeepAI.

Tupac Shakur's estate is none too happy about Drake cloning the late hip-hop legend's voice in a Kendrick Lamar diss track. Billboard reported Wednesday that attorney Howard King, representing Mr. Shakur's estate, sent a cease-and-desist letter calling Drake's use of Shakur's voice "a flagrant violation of Tupac's publicity and the estate's legal rights." Drake (Aubrey Drake Graham) dropped the diss track "Taylor Made Freestyle" last Friday, the latest chapter of the artist's simmering decade-long feud with Pulitzer and 17-time Grammy award winner Kendrick Lamar. "Kendrick, we need ya, the West Coast savior / Engraving your name in some hip-hop history," an AI-generated 2Pac recreation raps in Drake's track. "If you deal with this viciously / You seem a little nervous about all the publicity."

In the last week, highly anticipated songs by Drake and Taylor Swift appeared to leak online, sparking enormous reactions. Massive Reddit threads spawned, dissecting musical choices. Meme videos were created simulating other rappers' reactions to being dissed by Drake. The rapper Rick Ross even responded to the song's bars about him with a diss track of his own. But there was one big problem: neither Swift nor Drake confirmed that the songs were real.