Surveys have recently gained popularity as a tool to study large language models. By comparing models' survey responses to those of different human reference populations, researchers aim to infer the demographics, political opinions, or values best represented by current language models. In this work, we critically examine language models' survey responses on the basis of the well-established American Community Survey by the U.S. Census Bureau. Evaluating 43 different language models using de-facto standard prompting methodologies, we establish two dominant patterns. First, models' responses are governed by ordering and labeling biases, for example, towards survey responses labeled with the letter "A".

The accurate identification of active sites in proteins is essential for the advancement of life sciences and pharmaceutical development, as these sites are of critical importance for enzyme activity and drug design. Recent advancements in protein language models (PLMs), trained on extensive datasets of amino acid sequences, have significantly improved our understanding of proteins. However, compared to the abundant protein sequence data, functional annotations, especially precise per-residue annotations, are scarce, which limits the performance of PLMs. On the other hand, textual descriptions of proteins, which could be annotated by human experts or a pretrained protein sequence-to-text model, provide meaningful context that could assist in the functional annotations, such as the localization of active sites. This motivates us to construct a $\textbf{ProT}$ein-$\textbf{A}$ttribute text $\textbf{D}$ataset ($\textbf{ProTAD}$), comprising over 570,000 pairs of protein sequences and multi-attribute textual descriptions.

Plus: OpenAI is also creating a super app. OpenAI has a new grand challenge: building an AI researcher--a fully automated agent-based system capable of tackling large, complex problems by itself. The San Francisco firm said the new goal will be its "north star" for the next few years. By September, the company plans to build "an autonomous AI research intern" that can take on a small number of specific research problems. The intern will be the precursor to the fully automated multi-agent system, which is slated to debut in 2028. In an exclusive interview this week, OpenAI's chief scientist, Jakub Pachocki, talked me through the plans.

Photographer Tom Murphy has documented the park's wildlife for decades. Now, one of his photos will be on a Forever Stamp. The new stamp features one of Yellowstone's signature bison and will be out later in 2026. Breakthroughs, discoveries, and DIY tips sent six days a week. It's a warm July day in Yellowstone National Park's grassy Hayden Valley and wildlife photographer Tom Murphy is tracking herds of chocolate-colored bison gathered for the annual breeding season.

As part of the effort to understand how the brain learns, ongoing research seeks to combine biological knowledge with current artificial intelligence (AI) modeling in an attempt to find an efficient biologically plausible learning scheme. Current models often use a cortical-like combination of bottom-up (BU) and top-down (TD) processing, where the TD part carries feedback signals for learning. However, in the visual cortex, the TD pathway plays a second major role in visual attention, by guiding the visual process toward locations and tasks of interest. A biological model should therefore integrate both learning and visual guidance. We introduce a model that uses a cortical-like combination of BU and TD processing that naturally integrates the two major functions of the TD stream. This integration is achieved through an appropriate connectivity pattern between the BU and TD streams, a novel processing cycle that uses the TD stream twice, and a'Counter-Hebb' learning mechanism that operates across both streams. We show that the'Counter-Hebb' mechanism can provide an exact backpropagation synaptic modification. Additionally, our model can effectively guide the visual stream to perform a task of interest, achieving competitive performance on standard multi-task learning benchmarks compared to AI models. The successful combination of learning and visual guidance could provide a new view on combining BU and TD processing in human vision and suggests possible directions for both biologically plausible models and artificial instructed models, such as vision-language models (VLMs).

OpenAI is refocusing its research efforts and throwing its resources into a new grand challenge. The San Francisco firm has set its sights on building what it calls an AI researcher, a fully automated agent-based system that will be able to go off and tackle large, complex problems by itself. OpenAI says that this new research goal will be its "North Star" for the next few years, pulling together multiple research strands, including work on reasoning models, agents, and interpretability .

It filed a request with the FCC to deploy almost 52,000 satellites. Blue Origin has revealed its plans for an {@/data/467/1/1 orbital AI data center @/data/467/1/1} system in a new filing with the Federal Communications Commission. The company has asked the agency for permission to deploy 51,600 satellites, as reported by the and . Called Project Sunrise, the initiative aims to launch and operate a constellation of satellites that can deliver computing capacity for artificial intelligence uses. Project Sunrise's satellites will be placed in sun-synchronous orbits at altitudes between 311 and 1,118 miles.

Current adversarial attacks for multi-class classifiers choose potential adversarial target classes naively based on the classifier's confidence levels. We present a novel adversarial targeting method, \textit{MALT - Mesoscopic Almost Linearity Targeting}, based on local almost linearity assumptions. Our attack wins over the current state of the art AutoAttack on the standard benchmark datasets CIFAR-100 and Imagenet and for different robust models. In particular, our attack uses a \emph{five times faster} attack strategy than AutoAttack's while successfully matching AutoAttack's successes and attacking additional samples that were previously out of reach. We additionally prove formally and demonstrate empirically that our targeting method, although inspired by linear predictors, also applies to non-linear models.

While the United States and Europe are moving cautiously forward with clinical trials, China is racing toward the commercialization of brain implants. China has made history by becoming the first nation to approve a commercially available brain chip to treat a disability. NEO, the implant developed by Neuracle Medical Technology, translates the thoughts of a person with paralysis into movements of an assistive robotic hand. After 18 months of testing that proved its safety, China's National Medical Products Administration authorized the implant for people aged 19 to 60 with paralysis caused by neck or spinal cord injuries that prevent them from moving their limbs. According Nature, the implant embedded in the skull is about the size of a coin.

Generating ligand molecules for specific protein targets, known as structure-based drug design, is a fundamental problem in therapeutics development and biological discovery. Recently, target-aware generative models, especially diffusion models, have shown great promise in modeling protein-ligand interactions and generating candidate drugs. However, existing models primarily focus on learning the chemical distribution of all drug candidates, which lacks effective steerability on the chemical quality of model generations. In this paper, we propose a novel and general alignment framework to align pretrained target diffusion models with preferred functional properties, named AliDiff. AliDiff shifts the target-conditioned chemical distribution towards regions with higher binding affinity and structural rationality, specified by user-defined reward functions, via the preference optimization approach. To avoid the overfitting problem in common preference optimization objectives, we further develop an improved Exact Energy Preference Optimization method to yield an exact and efficient alignment of the diffusion models, and provide the closed-form expression for the converged distribution. Empirical studies on the CrossDocked2020 benchmark show that AliDiff can generate molecules with state-of-the-art binding energies with up to -7.07 Avg.