The widespread use of black box prediction methods has sparked an increasing interest in algorithm/model-agnostic approaches for quantifying goodness-of-fit, with direct ties to specification testing, model selection and variable importance assessment. A commonly used framework involves defining a predictiveness criterion, applying a cross-fitting procedure to estimate the predictiveness, and utilizing the difference in estimated predictiveness between two models as the test statistic. However, even after standardization, the test statistic typically fails to converge to a non-degenerate distribution under the null hypothesis of equal goodness, leading to what is known as the degeneracy issue. To addresses this degeneracy issue, we present a simple yet effective device, Zipper. It draws inspiration from the strategy of additional splitting of testing data, but encourages an overlap between two testing data splits in predictiveness evaluation. Zipper binds together the two overlapping splits using a slider parameter that controls the proportion of overlap. Our proposed test statistic follows an asymptotically normal distribution under the null hypothesis for any fixed slider value, guaranteeing valid size control while enhancing power by effective data reuse. Finite-sample experiments demonstrate that our procedure, with a simple choice of the slider, works well across a wide range of settings.

By stripping away the fabric tape that's held zippers together for a hundred years, Japanese clothing giant YKK is designing the future of seamless clothing. For more than a century, the zipper has stayed more or less the same: two interlocking rows of teeth, a sliding pull, and the fabric tape that holds it together. Billions are used every day, yet few people ever stop to think about how they work. Now, after a hundred years of stasis, YKK, the Japanese company that makes roughly half the world's zippers, has decided it's time to rethink the mechanism that holds much of modern clothing together. Their new AiryString zipper looks ordinary at first glance.

The widespread use of black box prediction methods has sparked an increasing interest in algorithm/model-agnostic approaches for quantifying goodness-of-fit, with direct ties to specification testing, model selection and variable importance assessment. A commonly used framework involves defining a predictiveness criterion, applying a cross-fitting procedure to estimate the predictiveness, and utilizing the difference in estimated predictiveness between two models as the test statistic. However, even after standardization, the test statistic typically fails to converge to a non-degenerate distribution under the null hypothesis of equal goodness, leading to what is known as the degeneracy issue. To addresses this degeneracy issue, we present a simple yet effective device, Zipper. It draws inspiration from the strategy of additional splitting of testing data, but encourages an overlap between two testing data splits in predictiveness evaluation.

For millennia, during Finland's blistering winters, wind drove snow into meters-high snowbanks along Lake Saimaa's shoreline, offering prime real estate from which seals carved cave-like dens to shelter from the elements and raise newborns. But in recent decades, these snowdrifts have failed to form in sufficient numbers, as climate change has brought warming temperatures and rain in place of snow, decimating the seal population. For the last 11 years, humans have stepped in to construct what nature can no longer reliably provide. Human-made snowdrifts, built using handheld snowplows, now house 90% of seal pups. They are the latest in a raft of measures that have brought Saimaa's seals back from the brink of extinction.

Integrating multiple generative foundation models, especially those trained on different modalities, into something greater than the sum of its parts poses significant challenges. Two key hurdles are the availability of aligned data (concepts that contain similar meaning but is expressed differently in different modalities), and effectively leveraging unimodal representations in cross-domain generative tasks, without compromising their original unimodal capabilities. We propose Zipper, a multi-tower decoder architecture that addresses these concerns by using cross-attention to flexibly compose multimodal generative models from independently pre-trained unimodal decoders. In our experiments fusing speech and text modalities, we show the proposed architecture performs very competitively in scenarios with limited aligned text-speech data. We also showcase the flexibility of our model to selectively maintain unimodal (e.g., text-to-text generation) generation performance by freezing the corresponding modal tower (e.g. text). In cross-modal tasks such as automatic speech recognition (ASR) where the output modality is text, we show that freezing the text backbone results in negligible performance degradation. In cross-modal tasks such as text-to-speech generation (TTS) where the output modality is speech, we show that using a pre-trained speech backbone results in superior performance to the baseline.

Questionable assertions are rapidly multiplying in the self-driving car space. Self-driving vehicles require relentless scrutiny. However, some fundamental misunderstandings cause industry observers to make broad definitive statements with very little to back up their conclusions. To respond each time would be one long game of Whack-A-Mole and there's only so much time in the day. But I'm motivated to push back after seeing the highly influential Washington Post devote a recent front page of their Outlook Section to an article titled "Companies Are Still Racing To Make Self-Driving Cars: But Why," with the subtitle "They may not be safer than human drivers.

It's not because Sophia has a zipper on the back of its head though it illustrates quite well how far we still are from an intelligent and human-like robot. Have you seen already the new attempt of Sophia The Robot and its manufacturer Hanson Robotics to make public believe this automaton has consciousness and intelligence? Just yesterday Sophia got into the news again, starring in a short movie'SophiaWorld' along with'Westworld' actress Evan Rachel Woods. Why it matters to understand that Sophia is not what Hanson Robotics tries to pretend it to be? Discrepancy of what AI is and isn't remains too big to fall for the show. Sophia The Robot does not even act up to its full name.

We revisit the SeqBin constraint. This meta-constraint subsumes a number of important global constraints like Change, Smooth and IncreasingNValue. We show that the previously proposed filtering algorithm for SeqBin has two drawbacks even under strong restrictions: it does not detect bounds disentailment and it is not idempotent. We identify the cause for these problems, and propose a new propagator that overcomes both issues. Our algorithm is based on a connection to the problem of finding a path of a given cost in a restricted $n$-partite graph. Our propagator enforces domain consistency in O(nd^2) and, for special cases of SeqBin that include Change, Smooth and IncreasingNValue, in O(nd) time.