Leveraging the power of Large Language Models (LLMs) through preference optimization is crucial for aligning model outputs with human values. Direct Preference Optimization (DPO) has recently emerged as a simple yet effective method by directly optimizing on preference data without the need for explicit reward models. However, DPO typically relies on human-labeled preference data, which can limit its scalability. Self-Play Fine-Tuning (SPIN) addresses this by allowing models to generate their own rejected samples, reducing the dependence on human annotations. Nevertheless, SPIN uniformly applies learning signals across all tokens, ignoring the fine-grained quality variations within responses. As the model improves, rejected samples increasingly contain high-quality tokens, making the uniform treatment of tokens suboptimal. In this paper, we propose SWIFT (Self-Play Weighted Fine-Tuning), a fine-grained self-refinement method that assigns token-level importance weights estimated from a stronger teacher model. Beyond alignment, we also demonstrate that SWIFT serves as an effective knowledge distillation strategy by using the teacher not for logits matching, but for reward-guided token weighting. Extensive experiments on diverse benchmarks and settings demonstrate that SWIFT consistently surpasses both existing alignment approaches and conventional knowledge distillation methods.

Researchers show scammers are using AI-manipulated footage of celebrity interviews to trick users into sharing their personal data. Last week, Taylor Swift filed a trio of trademark applications to protect her image and voice. One is meant to cover a well-known photograph of the pop singer holding a pink guitar during a concert on her record-breaking Eras tour, while the two sound trademarks are for simple identifying phrases: "Hey, it's Taylor Swift" and "Hey, it's Taylor." The move comes as AI deepfakes continue to proliferate across social media. Any individual stands to have their likeness exploited in the creation of nonconsensual AI-generated material; earlier this month, an Ohio man was the first person convicted under a new federal law criminalizing "intimate" visual deceptions of this sort.

Taylor Swift has applied to trademark her voice and appearance in an apparent attempt to protect herself from artificial intelligence impersonations. The pop superstar has lodged three trademark applications in the US - one using a photo of herself on stage during her Eras Tour, and the other two being audio clips of her introducing herself while promoting her last album. AI-generated versions of Swift have cropped up in various ways in recent years - from explicit images to a fake election ad in which she appeared to urge people to vote for Donald Trump. The move comes after actor Matthew McConaughey became the first celebrity to use trademark rules to attempt to protect his voice and image from AI misuse earlier this year . Trademark applications are a relatively new way for celebrities to combat the growing issue of AI rip-offs.

Modeling spatiotemporal brain dynamics from high-dimensional data, such as functional Magnetic Resonance Imaging (fMRI), is a formidable task in neuroscience.

Modeling spatiotemporal brain dynamics from high-dimensional data, such as functional Magnetic Resonance Imaging (fMRI), is a formidable task in neuroscience. Existing approaches for fMRI analysis utilize hand-crafted features, but the process of feature extraction risks losing essential information in fMRI scans. To address this challenge, we present SwiFT (Swin 4D fMRI Transformer), a Swin Transformer architecture that can learn brain dynamics directly from fMRI volumes in a memory and computation-efficient manner. SwiFT achieves this by implementing a 4D window multi-head self-attention mechanism and absolute positional embeddings. We evaluate SwiFT using multiple large-scale resting-state fMRI datasets, including the Human Connectome Project (HCP), Adolescent Brain Cognitive Development (ABCD), and UK Biobank (UKB) datasets, to predict sex, age, and cognitive intelligence.

Taylor Swift fans are driving a surge in popularity of a German museum exhibiting a portrait of the Shakespeare character Ophelia, recently reimagined in a song and music video from Swift's new album The Life of a Showgirl. The Hessische Landesmuseum in the central German city of Wiesbaden saw hundreds more visitors than usual over the weekend, as fans hoped to see the real version of the painting that opens the music video for The Fate of Ophelia. In the video, viewed more than 65 million times on Youtube, the painting comes alive, with Swift at its centre. We're really enjoying this attention - it's a lot of fun, museum spokesperson Susanne Hirschmann told the Associated Press. Hirschmann said that one family had travelled from the northern city of Hamburg, a five-hour drive away, while some of the visitors were Americans from an army base nearby.

Fans Call on Taylor Swift to'Do Better' After Accusations of Using AI for Promo Videos A scavenger hunt campaign to promote Taylor Swift's new album, resulted in a viral #SwiftiesAgainstAI campaign. Fans attend a screening of at a theater in Los Angeles. These were just some of the alleged clues that fans spotted in promo videos for Taylor Swift's new album,, this weekend. They were, to their eyes, telltale indicators that the videos were purportedly made with generative AI . "The first sign that it was AI was that it didn't look great," claims Marcela Lobo, a graphic designer in Brazil who has been a Swift fan since she was 12. "It was wonky, the shadows didn't match, the windows and the painted piano, it looked like shit, basically."

Charlie Kirk leaked text confirms he was livid about'bullying' Jewish donors: 'I'm leaving pro-Israel cause' White House insider who says WAR with Venezuela is inevitable... as Trump's lethal options are laid out I've seen the real Victoria Beckham... her actions gave me PTSD, she shunned me and even banned me from glancing in her direction. Jimmy Kimmel's audience boom comes crashing down as he loses 71% of viewers in one week'Kissing Trump's a**': President mocks Canada's obsequious PM as he begs for tariff relief World's most invasive predator terrorizing East Coast is delicious and should be eaten to stop its spread, experts say I've had enough of the arrogant and entitled fat brigade. Bloodcurdling videos shows girl aged 12 subway surfing days before she and friend, 13, died during 3.10am stunt Another blow for Prince Harry as African country cuts ties with his'disrespectful' charity Friends fear for new CBS News boss Bari Weiss, claiming her wife thinks she sold out... and her new job will'consume her life' Keith Urban's guitarist Maggie once vowed to'never' date a tour mate... as she's accused of charming Nicole Kidman's ex Hollywood's favorite muscle car primed for return as America's No.1 automaker files secret paperwork AMANDA PLATELL: I never thought I'd feel sorry for Harry. There's one thing he'd do anything to defend... and now Meghan's trampled all over it Ben Affleck's VERY familiar whispers to Jennifer Lopez on the red carpet revealed... as their romantic new era sends fans into overdrive Jimmy Kimmel continues anti-Trump rants and says he's more popular with Americans than the president Brits have been left in awe after spotting what is believed to be a'meteor' glowing through the night sky. Lucky stargazers in Northfields and West Ealing, west London, have reported seeing a blue-ish green blob race through the city's sky tonight.

Diffusion models offer a physically grounded framework for probabilistic weather forecasting, but their typical reliance on slow, iterative solvers during inference makes them impractical for subseasonal-to-seasonal (S2S) applications where long lead-times and domain-driven calibration are essential. To address this, we introduce Swift, a single-step consistency model that, for the first time, enables autoregressive finetuning of a probability flow model with a continuous ranked probability score (CRPS) objective. This eliminates the need for multi-model ensembling or parameter perturbations. Results show that Swift produces skillful 6-hourly forecasts that remain stable for up to 75 days, running $39\times$ faster than state-of-the-art diffusion baselines while achieving forecast skill competitive with the numerical-based, operational IFS ENS. This marks a step toward efficient and reliable ensemble forecasting from medium-range to seasonal-scales.