Vision-language alignment in Large Vision-Language Models (LVLMs) successfully enables LLMs to understand visual input. However, we find that existing vision-language alignment methods fail to transfer the existing safety mechanism for text in LLMs to vision, which leads to vulnerabilities in toxic image. To explore the cause of this problem, we give the insightful explanation of where and how the safety mechanism of LVLMs operates and conduct comparative analysis between text and vision. We find that the hidden states at the specific transformer layers play a crucial role in the successful activation of safety mechanism, while the vision-language alignment at hidden states level in current methods is insufficient. This results in a semantic shift for input images compared to text in hidden states, therefore misleads the safety mechanism. To address this, we propose a novel Text-Guided vision-language Alignment method (TGA) for LVLMs. TGA retrieves the texts related to input vision and uses them to guide the projection of vision into the hidden states space in LLMs. Experiments show that TGA not only successfully transfers the safety mechanism for text in basic LLMs to vision in vision-language alignment for LVLMs without any safety fine-tuning on the visual modality but also maintains the general performance on various vision tasks (Safe and Good). Vision-language alignment methods for Large Vision-Language Models (LVLMs) use a basic LLM, a lightweight vision encoder and projector to efficiently enable the LLM to understand visual input for various vision tasks with relatively low training costs (Liu et al., 2024c; Dai et al., 2023; Zhu et al., 2023). Recent studies indicate the safety of LVLMs deserves attention (Liu et al., 2024a; Wang et al., 2023; Gong et al., 2023). Given that vision and language are aligned into a common space in LVLMs, the safety mechanism should be shared by both of them.

A senior California lawmaker will introduce a new artificial intelligence (AI) bill to the state's senate on Wednesday, adding to national and global efforts to regulate the fast-accelerating technology. Although there are several attempts in Congress to draft AI legislation, the state of California--home to Silicon Valley, where most of the world's top AI companies are based--has a role to play in setting guardrails on the industry, according to state Senator Scott Wiener, (D--San Francisco) who drafted the bill. "In an ideal world we would have a strong federal AI regulatory scheme," Wiener said in an interview with TIME on Tuesday, adding that he supports attempts in Congress and the White House to regulate the technology. "But California has a history of acting when the federal government is moving either too slowly or not acting." He added: "We need to get ahead of these risks, not do what we've done in the past around social media or other technology, where we do nothing before it's potentially too late."

For robots to successfully execute tasks assigned to them, they must be capable of planning the right sequence of actions. These actions must be both optimal with respect to a specified objective and satisfy whatever constraints exist in their world. We propose an approach for robot task planning that is capable of planning the optimal sequence of grounded actions to accomplish a task given a specific objective function while satisfying all specified numerical constraints. Our approach accomplishes this by encoding the entire task planning problem as a single mixed integer convex program, which it then solves using an off-the-shelf Mixed Integer Programming solver. We evaluate our approach on several mobile manipulation tasks in both simulation and on a physical humanoid robot. Our approach is able to consistently produce optimal plans while accounting for all specified numerical constraints in the mobile manipulation tasks. Open-source implementations of the components of our approach as well as videos of robots executing planned grounded actions in both simulation and the physical world can be found at this url: https://adubredu.github.io/gtpmip

Texas border town barbeque owner says she was forced to sell her family business after migrants broke in and stole registers, computers and family heirlooms. Gov. Greg Abbott invoking the U.S. and Texas constitutions' invasion clauses won't change anything if illegal immigrants aren't deported, Ken Cuccinelli, a former Trump administration official, told Fox News. Abbott announced Tuesday that he invoked the clauses to "defend our state against an invasion" to combat the record-setting wave of illegal immigration occurring along the border. The Texas Republican sent a letter to Texas county judges indicating that he would deploy the National Guard and the Texas Department of Public Safety and build a border wall, among other actions. "If the goal is actually to reduce the harm to Texas, then he needs to start using the invasion authority to return people to Mexico," said Ken Cuccinelli, who served as a Department of Homeland Security acting deputy secretary under former President Trump.

The current pandemic has introduced substantial uncertainty to traditional methods for demand planning. These uncertainties stem from the disease progression, government interventions, economy and consumer behavior. While most of the emerging literature on the pandemic has focused on disease progression, a few have focused on consequent regulations and their impact on individual behavior. The contributions of this paper include a quantitative behavior model of fear of COVID-19, impact of government interventions on consumer behavior, and impact of consumer behavior on consumer choice and hence demand for goods. It brings together multiple models for disease progression, consumer behavior and demand estimation-thus bridging the gap between disease progression and consumer demand. We use panel regression to understand the drivers of demand during the pandemic and Bayesian inference to simplify the regulation landscape that can help build scenarios for resilient demand planning. We illustrate this resilient demand planning model using a specific example of gas retailing. We find that demand is sensitive to fear of COVID-19: as the number of COVID-19 cases increase over the previous week, the demand for gas decreases -- though this dissipates over time. Further, government regulations restrict access to different services, thereby reducing mobility, which in itself reduces demand.

State officials hope California's new 10 p.m. stay-at-home order will slow the spread of COVID-19, otherwise, another 10,000 San Diegans are projected to contract the virus in the next 10 days. That's according to a new county-by-county forecast from Facebook, which rolled out the prediction software last month. Facebook projects L.A. County will see the second-largest increase in cases in the country by November 30. San Diego County is projected to add the 15th most cases, reaching a total of 78,594 infections by Nov. 30. The two-week forecast was released before Governor Gavin Newsom announced enhanced restrictions.

With influenza cases elevated nationally and widespread throughout the country, researchers led by Boston Children's Hospital contend that machine learning can produce highly accurate local flu surveillance. In fact, they say that combining two forecasting methods with artificial intelligence produces the most accurate estimates of flu activity available to date--a week ahead of traditional healthcare-based reports, at the state level across the United States. While the Centers for Disease Control and Prevention monitors influenza-like illnesses (ILI) in the U.S. by gathering information from physicians' reports about patients with ILI seeking medical attention, the availability of the data has a lag time of as much as two weeks. However, in a study published on Friday in Nature Communications, researchers say they have successfully combined Google search frequencies and electronic health record data with spatio-temporal trends in influenza activity to produce forecasts with higher correlation and lower errors than all other tested models for current ILI activity at the state level. "We believe that the accuracy of our method involves a balance between responsiveness and robustness," state the authors.