Offline-to-online learning aims to improve online decision-making by leveraging offline logged data. A central challenge in this setting is the distribution shift between offline and online environments. While some existing works attempt to leverage shifted offline data, they largely rely on UCB-type algorithms. Thompson sampling (TS) represents another canonical class of bandit algorithms, well known for its strong empirical performance and naturally suited to offline-to-online learning through its Bayesian formulation. However, unlike UCB indices, posterior samples in TS are not guaranteed to be optimistic with respect to the true arm means. This makes indices constructed from purely online and hybrid data difficult to compare and complicates their use. To address this issue, we propose sample-mean anchored TS (Anchor-TS), which introduces a novel median-based anchoring rule that defines the arm index as the median of an online posterior sample, a hybrid posterior sample, and the online sample mean. The median anchoring systematically corrects bias induced by distribution shift by mitigating over-estimation for suboptimal arms and under-estimation for optimal arms, while exploiting offline information to obtain more accurate estimates when the shift is small. We establish theoretical guarantees showing that the proposed algorithm safely leverages offline data to accelerate online learning, and quantifying how the degree of distribution shift and the size of offline data affect the resulting regret reduction. Extensive experiments demonstrate consistent improvements of our algorithm over baselines.

Supervised fine-tuning (SFT) provides the standard approach for teaching LLMs new behaviors from offline expert demonstrations. However, standard SFT uniformly fits all samples -- including those with low likelihood under the base model -- which can disproportionately drive training updates toward overfitting specific samples rather than learning the target behavior. Moreover, adapting to these unlikely samples induces substantial policy shifts that degrade prior capabilities. Existing methods mitigate this by filtering, regenerating, or down-weighting low-likelihood data. In doing so, they often suppress precisely the novel behaviors the base model has yet to learn. We propose InfoSFT, a principled weighting scheme for the SFT objective that concentrates learning signals on maximally informative, medium-confidence tokens -- those neither overly familiar to the base model nor too unlikely to cause instability. Requiring only a one-line modification to the standard token-wise loss, InfoSFT demonstrably improves generalization over vanilla SFT and likelihood-weighted baselines across math, code, and chain-of-thought tasks with diverse model families, while better preserving pre-existing capabilities.

Valiant's 1984 paper is widely credited with introducing the PAC learning model, but it, in fact, introduced a different model: unlike PAC learning, the learner receives only positives, may issue membership queries, and must output a hypothesis with no false positives. Prior work characterized variants, including the case without queries. We revisit Valiant's original model and ask: *Which classes are learnable in it?* For every finite domain, including Valiant's Boolean-hypercube setting, we show that a class is learnable if and only if every realizable positive sample can be certified by a poly-size adaptive query-compression scheme. This is a new variant of sample compression where the learner certifies samples via a short interaction with the membership oracle. Our characterization shows that learnability in Valiant's model is strictly sandwiched between learnability in the PAC model and the variant of Valiant's model without membership queries. This is one of the rare cases where introducing membership queries changes the set of learnable classes, and not just the sample or computational complexity. Next, we study the natural extension of the model to arbitrary domains. While we do not obtain an exact characterization, our techniques readily generalize and show that the same strict sandwiching persists. Finally, we show that $d$-dimensional halfspaces, which are not learnable without queries, are learnable with queries: we give a $\mathrm{poly}(d) \tilde{O}(1/ε)$ sample and $\mathrm{poly}(d) \mathrm{polylog}(1/ε)$ query algorithm, and prove that at least $Ω(d)$ samples or queries are necessary. To our knowledge, this is the first algorithm for halfspaces in Valiant's model. Together, these results uncover a surprisingly rich theory behind Valiant's original notion of learnability and introduce ideas that may be of independent interest in learning theory.

The 10th International Workshop on Health Intelligence (W3PHIAI-26) celebrated a decade of bringing AI and health research together, building on a lineage that began with the AAAI-W3PHI workshops focused on population health (2014-2016), the AAAI-HIAI workshops focused on personalized health (2013-2016), and the subsequent joint W3PHIAI workshops held annually from 2017 through 2025. Over this decade, the series has produced hundreds of talks and high-impact publications that have collectively received thousands of citations, shaping the research agenda in both population health intelligence and personalized healthcare AI. This year's special theme, "Foundation Models and AI Agents," reflected the field's rapidly evolving frontier: the emergence of autonomous and semi-autonomous AI systems reshaping clinical workflows, patient management, health system operations, and public health surveillance. Day 1 of the workshop focused on medical imaging and the translation of AI for clinical ...

Realtor's evil ex-husband avoids murder trial with guilty plea after'ambushing' mother of two while she begged for her life on 911 call Explosive Supreme Court LEAK reveals stinging whispers about'belligerent' justice read the wild rants troubling both sides of the aisle Married doctor's affair with glamorous younger woman explodes into Fatal Attraction-style court war... X-rated photo claims, leaked recordings and a sinister threat: 'I'll never stop' US intel reportedly says Iran's military is FAR from decimated as Israel begins to worry about Trump's deal-making Michelle Obama looks alarmingly thin on Beverly Hills dinner date with Malia and Sasha - as Barack's absence fuels fresh whispers about their marriage Brady Bunch's Eve Plumb reveals cast's shocking residual pay after Lisa Kudrow said Friends stars still get $20m a year from reruns The unassuming apps all cheaters use to hide their affairs: Where to look on your partner's phone to see exactly what they are up to... and the subtle red flags to never ignore I've treated so many cocaine users. This is the one sign that makes it so obvious you have a problem, how it can kill you in a night... and the embarrassing sexual side effect you may not have heard of: DR PHILIPPA KAYE Demi Moore, 63, sparks concern with thinner-than-ever frame at Cannes... amid swirling Ozempic rumors High school student singles out board member who called her'hot' with humiliating takedown Lindsay Lohan, 39, baffles fans with'unrecognizable' appearance at Disney Upfronts event reigniting plastic surgery rumors The'marry me' sex move that'll make even the most commitment-phobic of men beg to see you again... and it worked for THREE of my friends Real-life Pacific Rim! World's first manned transforming robot is unveiled - and it could be yours for $650,000 A Chinese robotics firm has truly bridged the gap between science fiction and reality, after unveiling the world's first manned'mecha'. The GD01, developed by Unitree Robotics, weighs 500kg with a pilot on board and is capable of transitioning between bipedal walking and four-legged mode. Developed for civilian transport, the high-strength alloy machine features a'cockpit' where someone can sit and control the huge robot. A demonstration video shows Unitree's CEO Wang Xingxing climbing into the torso of the GD01 before it starts to move.

We study the information-theoretic limits of learning a one-hidden-layer teacher network with hierarchical features from noisy queries, in the context of knowledge transfer to a smaller student model. We work in the high-dimensional regime where the teacher width $k$ scales linearly with the input dimension $d$ -- a setting that captures large-but-finite-width networks and has only recently become analytically tractable. Using a heuristic leave-one-out decoupling argument, validated numerically throughout, we derive asymptotically sharp characterizations of the Bayes-optimal generalization error and individual feature overlaps via a system of closed fixed-point equations. These equations reveal that feature learnability is governed by a sequence of sharp phase transitions: as data grows, teacher features become recoverable sequentially, each through a discontinuous jump in overlap. This sequential acquisition underlies a precise notion of \textit{effective width} $k_c$ -- the number of learnable features at a given data budget $n$ -- which unifies two distinct scaling regimes: a feature-learning regime in which the Bayes-optimal generalization error $\varepsilon^{\rm BO}$ scales as $ n^{1/(2β)-1}$, and a refinement regime in which it scales as $n^{-1}$, where $β>1/2$ is the exponent of the power-law feature hierarchy. Both laws collapse to the single relation $\varepsilon^{\rm BO}=Θ(k_c d/n)$. We further show empirically that a student trained with \textsc{Adam} near the effective width $k_c$ achieves these optimal scaling laws (up to a small algorithmic gap), and provide an information-theoretic account of the associated scaling in model size.

The trustworthiness of AI decision-making systems is increasingly important. A key feature of such systems is the ability to provide recommendations for how an individual may reverse a negative decision, a problem known as algorithmic recourse. Existing approaches treat recourse outcomes as counterfactuals of a fixed unit, ignoring that real-world recourse involves repeated decisions on the same individual under possibly different latent conditions. We develop a causal framework that models recourse as a process over pre- and post-intervention outcomes, allowing for partial stability and resampling of latent variables. We introduce post-recourse stability conditions that enable reasoning about recourse from observational data alone, and develop a copula-based algorithm for inferring the effects of recourse under these conditions. For settings where paired observations of the same individual before and after intervention are available (called recourse data), we develop methods for inferring copula parameters and performing goodness-of-fit testing. When the copula model is rejected, we provide a distribution-free algorithm for learning recourse effects directly from recourse data. We demonstrate the value of the proposed methods on real and semi-synthetic datasets.

Neural operators provide a framework for learning solution operators of partial differential equations (PDEs), enabling efficient surrogate modeling for complex systems. While universal approximation results are now well understood, approximation analysis specific to nonlinear reaction-diffusion systems remains limited. In this paper, we study neural operators applied to the solution mapping from initial conditions to time-dependent solutions of a generalized Gierer-Meinhardt reaction-diffusion system, a prototypical model of nonlinear pattern formation. Our main results establish explicit approximation error bounds in terms of network depth, width, and spectral rank by exploiting the Laplacian spectral representation of the Green's function underlying the PDE. We show that the required parameter complexity grows at most polynomially with respect to the target accuracy, demonstrating that Laplacian eigenfunction-based neural operator architectures alleviate the curse of parametric complexity encountered in generic operator learning. Numerical experiments on the Gierer-Meinhardt system support the theoretical findings.

Hang gliding Lookout Mountain: What it's really like to be aero-towed 1,700 feet above Georgia Paige Spiranac and her mom stun the internet, Lane Kiffin's incredible shot at Ole Miss & the NFL did it again Maggie Sajak appears at Savannah Bananas game as Jackson Olson's girlfriend, e-bike near death & MEAT! Pulling a gun on your sister's boyfriend and telling him to strip is one way to make family events awkward Paige Spiranac hits bombs at Truist pro-am after years of being shunned, fighter jets interrupt golf & MEAT! Spencer Pratt is'channeling the frustration' of LA voters, 'Ruthless' co-host says'High degree of coordination': Calif mayor admits to being secret Chinese agent Trump predicts'a lot of good things' will happen in China, predicts'golden age of America' after Iran conflict ends Trump says Iran ceasefire on'massive life support' amid Middle East tensions Iran ceasefire on'massive life support' as Trump weighs military options Graham calls out China link to Iran, questions Pakistan's role in negotiations Caine accuses Iran of holding'world's economy hostage' with Strait of Hormuz actions'Fox & Friends' hosts learn backyard camping tips from Scouting America OutKick UCF graduates clobber commencement speaker with boos after she says AI is the'next Industrial Revolution' Gloria Caulfield addressed graduates from UCF's College of Arts and Humanities and the Nicholson School of Communication It's almost the end of the school year, which means it's graduation season, a time where commencement speakers across the nation will be giving boilerplate advice to hungover students. However, one speaker at the University of Central Florida, my alma mater -- Go Knights, Charge On! -- had a rough day at the podium thanks to a comment she made about artificial intelligence . According to Orlando Weekly, UCF held a graduation ceremony for the school's College of Arts and Humanities and the Nicholson School of Communication and Media last week, and the commencement speaker was vice president of strategic alliances for Tavistock Development Company, Gloria Caulfield.

Teen builds'Bionic Underwater Robotic Turtle' to detect ecological threats High schooler Evan Budz's award-winning invention can identify coral bleaching, invasive species, and microplastics without disturbing marine ecosystems. More information Adding us as a Preferred Source in Google by using this link indicates that you would like to see more of our content in Google News results. Canadian high school student Evan Budz poses with his award-winning bionic turtle. Breakthroughs, discoveries, and DIY tips sent six days a week. Fifteen-year-old Evan Budz was on a camping trip when he saw a snapping turtle that would become the impetus for an award-winning invention .