The latest developments in DNA sequencing techniques have revealed genes that play a role in determining our vulnerability to diseases and have given us a deeper understanding of our genetic composition. A groundbreaking milestone in genetic engineering has transformed the capabilities of the scientific community in altering the genetic material of different organisms. Among recent innovations, the clustered regularly interspaced short palindromic repeats (CRISPR) associated protein 9 (Cas9) has emerged as a powerful and precise tool for genome editing across diverse organisms. Its applications span immunotherapy, agriculture, poultry science, and human therapeutics, marking a transformative shift in biomedical and biotechnological research. However, the rapid progress and clinical translation of CRISPR/Cas9 have raised significant concerns regarding off-target effects, delivery challenges, long-term safety, and ethical implications. This review critically evaluates the CRISPR/Cas9 system by examining its molecular mechanism, editing efficiency, gene delivery approaches, and potential for inducing unintended mutations. A comparative analysis with other gene-editing tools is presented, emphasizing the advantages of CRISPR/Cas9 in programmability and editing efficiency. Furthermore, we discuss current advances including base editing, prime editing, and high-fidelity Cas variants, along with the ethical and societal dimensions of genome editing. This synthesis provides an updated perspective on the potential and limitations of CRISPR/Cas9 technology and highlights key areas for future research and responsible application.
Research in artificial intelligence is shifting from model innovations and benchmark scores towards problem definition and rigorous real-world evaluation. As the field enters the "second half," the central challenge becomes real utility in long-horizon, dynamic, and user-dependent settings such as agentic coding, deep research, and computer use, where LLM-based agents face context explosion beyond fixed context windows and must continuously accumulate, manage, and selectively reuse information across extended interactions. Memory, with hundreds of papers released in 2025, therefore emerges as the critical solution to fill this utility gap. Beyond passive storage, memory is increasingly the substrate through which agents self-evolve: short-term memory gates which experiences are perceived and abstracted during execution, while long-term memory consolidates them into reusable knowledge and skills, forming the loop through which agents improve from their own experience. In this survey, we provide a unified view of foundation agent memory along three dimensions: memory substrate (internal parametric state and external retrieval-augmented stores), cognitive mechanism (sensory, working, episodic, semantic, and procedural), and memory subject (user-centric personalization and agent-centric experience). We then analyze how memory is operated under single- and multi-agent topologies and highlight learning policies over memory operations, showing how memory management itself is becoming a trainable capability spanning reinforcement-learned context curation, experience consolidation at decision time, and the emerging ecosystem of portable, shareable agent skills. Finally, we review evaluation benchmarks and metrics for memory utility, and outline open challenges and future directions.
Mixture-of-Experts (MoE) architectures are increasingly used to efficiently scale large language models. However, in production inference, request batching and speculative decoding significantly amplify expert activation, eroding these efficiency benefits. We address this issue by modeling batch-aware expert selection as a modular optimization problem and designing efficient greedy algorithms for different deployment settings. The proposed method, namely XShare, requires no retraining and dynamically adapts to each batch by maximizing the total gating score of selected experts. It reduces expert activation by up to 30
Articulation modeling enables robots to learn joint parameters of articulated objects for effective manipulation which can then be used downstream for skill learning or planning. Existing approaches often rely on prior knowledge about the objects, such as the number or type of joints. Some of these approaches also fail to recover occluded joints that are only revealed during interaction. Others require large numbers of multi-view images for every object, which is impractical in real-world settings. Furthermore, prior works neglect the order of manipulations, which is essential for many multi-DoF objects where one joint must be operated before another, such as a dishwasher. We introduce PokeNet, an end-to-end framework that estimates articulation models from a single human demonstration without prior object knowledge. Given a sequence of point cloud observations of a human manipulating an unknown object, PokeNet predicts joint parameters, infers manipulation order, and tracks joint states over time. PokeNet outperforms existing state-of-the art methods, improving joint axis and state estimation accuracy by an average of over 27% across diverse objects, including novel and unseen categories. We demonstrate these gains in both simulation and real-world environments.
The Habitable Worlds Observatory (HWO) will require scalable detector readout architectures capable of supporting large-format Microwave Kinetic Inductance Detector (MKID) arrays with high tone density and stringent spectral isolation. Polyphase filterbank (PFB) channelizers have been widely adopted in MKID readouts on ground and suborbital platforms. However, translating these readouts to a space-qualified FPGA environment introduces additional constraints on determinism, traceability, verification, and resource utilization. This work presents a resource-efficient, fixed-point implementation of a critically sampled PFB coarse channelizer targeting space-readout development. The architecture is validated through FPGA hardware measurements and provides a traceable foundation for a future hand-coded VHDL implementation suitable for NASA review and certification.
