The maker of Ozempic and Wegovy is suing a rival firm for selling what it says are unsafe, knock-off versions of its weight-loss drugs in the US. Danish company Novo Nordisk asked US courts on Monday to ban Hims & Hers' range of weight-loss pills and injections, which it says are not approved by US authorities and infringe on its patent. The legal drama began on Friday after Hims & Hers launched a new weight-loss pill, leading to an initial threat from Novo Nordisk. Over the weekend, Hims & Hers said it would stop selling the pill. On Monday, its share price slumped as it called Novo Nordisk's decision to press ahead with the lawsuit a blatant attack.

Vision-and-Language Navigation (VLN) is a challenging task that requires an agent to navigate through complex environments based on natural language instructions. In contrast to conventional approaches, which primarily focus on the spatial domain exploration, we propose a paradigm shift toward the Fourier domain. This alternative perspective aims to enhance visual-textual matching, ultimately improving the agent's ability to understand and execute navigation tasks based on the given instructions. In this study, we first explore the significance of high-frequency information in VLN and provide evidence that it is instrumental in bolstering visual-textual matching processes. Building upon this insight, we further propose a sophisticated and versatile Frequency-enhanced Data Augmentation (FDA) technique to improve the VLN model's capability of capturing critical high-frequency information. Specifically, this approach requires the agent to navigate in environments where only a subset of high-frequency visual information corresponds with the provided textual instructions, ultimately fostering the agent's ability to selectively discern and capture pertinent high-frequency features according to the given instructions. Promising results on R2R, RxR, CVDN and REVERIE demonstrate that our FDA can be readily integrated with existing VLN approaches, improving performance without adding extra parameters, and keeping models simple and efficient. The code is available at https://github.com/hekj/FDA.

T o address the trade-off between robustness and performance for robust VLM, we observe that function words could incur vulnerability of VLMs against cross-modal adversarial attacks, and propose Function-word De-Attention (FDA) accordingly to mitigate the impact of function words. Similar to differential amplifiers, our FDA calculates the original and the function-word cross-attention within attention heads, and differentially subtracts the latter from the former for more aligned and robust VLMs. Comprehensive experiments include 2 SOTA baselines under 6 different attacks on 2 downstream tasks, 3 datasets, and 3 models. Overall, our FDA yields an average 18/13/53% ASR drop with only 0.2/0.3/0.6% performance drops on the 3 tested models on retrieval, and a 90% ASR drop with a 0.3% performance gain on visual grounding. W e demonstrate the scalability, generalization, and zero-shot performance of FDA experimentally, as well as in-depth ablation studies and analysis. Code will be made publicly available.

Figure 5: Layer notation of whitebox models and sequence in which layers get added to multi-intermediate-layer attacks. Here we discuss the DNN layer notation used throughout the work. Of the 16 possible layers we notate 14 of them in Figure 5 where "deeper" layers closer to the output of the model have higher layer numbers. The "Sequence" column is the order in which the 5 We use this exact architecture regardless of layer and feature map size, for simplicity reasons. We adopt a similar scheme here.

"straightforward" from simply looking at the equations, we maintain that the multi-layer extension is a significant However, note from Figure 5 (appendix) the pattern in which the layers are sequentially "added" by the We consider the direction of finding other optimizations for layer choice an important future work. From eqn 3, you are correct, it is possible for all layers to contribute differently. Intuitively, the most impactful layers are added first. The decoding for this layer notation is shown in Figure 5 (appendix). We will be sure to clarify these points in the final version.

We thank all reviewers for careful review and comments. This "sensitivity" is quantified in the Lipschitz constants For the toy example used to generate Figure 1 and Table 1 (described in Section 4.1), we know that the true Another approach is to project away the null space as we discussed in Line 216-218. The definition was moved to the supplementary material. When λ is small, WDA focuses more on global information and is more similar to FDA. In the context of classical FDA, these are two definitions that are commonly used in the literature.

Model merging is an efficient post-training strategy for integrating knowledge from multiple finetuned checkpoints of a shared foundation model. Existing methods operate in the parameter space, combining task vectors to mitigate conflicts, but remain constrained by parameter inconsistencies. We propose Functional Dual Anchors (FDAs), a framework that instead models the input-representation space. FDAs are synthetic inputs whose induced gradients align with task vectors, capturing task-specific functional shifts relative to the pretrained model. We further introduce a principled initialization scheme and show that FDAs are complementary to parameter-space model merging. Comprehensive experiments demonstrate the effectiveness of FDAs in model merging. Model merging has emerged as a promising post-training strategy for integrating knowledge from multiple finetuned checkpoints of foundation models. The core idea is to combine diverse domain knowledge from multiple homologous downstream models into a single unified one (Matena & Raffel, 2022; Jin et al., 2022). Compared to multi-task learning (Ruder, 2017) and continual learning (Wang et al., 2024), model merging is appealing because it consolidates knowledge directly through the parameters of downstream models finetuned from the same pretrained backbone. On the left, we compare multi-task joint training, task arithmetic and FDA. Inspired by joint training, FDA models the knowledge in the input space.

Large language models readily memorize arbitrary training instances, such as label noise, yet they perform strikingly well on reasoning tasks. In this work, we investigate how language models memorize label noise, and why such memorization in many cases does not heavily affect generalizable reasoning capabilities. Using two controllable synthetic reasoning datasets with noisy labels, four-digit addition (FDA) and two-hop relational reasoning (THR), we discover a reliance of memorization on generalizable reasoning mechanisms: models continue to compute intermediate reasoning outputs even when retrieving memorized noisy labels, and intervening reasoning adversely affects memorization. We further show that memorization operates through distributed encoding, i.e., aggregating various inputs and intermediate results, rather than building a look-up mechanism from inputs to noisy labels. Moreover, our FDA case study reveals memorization occurs via outlier heuristics, where existing neuron activation patterns are slightly shifted to fit noisy labels. Together, our findings suggest that memorization of label noise in language models builds on, rather than overrides, the underlying reasoning mechanisms, shedding lights on the intriguing phenomenon of benign memorization.

Speaking at a WIRED event Tuesday, Moderna CEO Stéphane Bancel said he was "encouraged" by the company's dialogue with the FDA--but acknowledged recent setbacks. Moderna CEO Stéphane Bancel prepares to testify before the Senate on March 22, 2023 in Washington, DC. At the WIRED Health summit on Tuesday, Moderna CEO Stéphane Bancel said the recent changes to Covid-19 vaccine policy made by Health and Human Services secretary Robert F. Kennedy, Jr. are a "step backward." Moderna is one of the manufacturers of mRNA-based Covid-19 vaccines, and last month the company received approval from the Food and Drug Administration for an updated version of the shot . But as part of that approval, the FDA imposed new restrictions on who can receive the vaccine.