Large language models (LLMs) are having transformative impacts across a wide range of scientific fields, particularly in the biomedical sciences. Just as the goal of Natural Language Processing is to understand sequences of words, a major objective in biology is to understand biological sequences. Genomic Language Models (gLMs), which are LLMs trained on DNA sequences, have the potential to significantly advance our understanding of genomes and how DNA elements at various scales interact to give rise to complex functions. In this review, we showcase this potential by highlighting key applications of gLMs, including fitness prediction, sequence design, and transfer learning. Despite notable recent progress, however, developing effective and efficient gLMs presents numerous challenges, especially for species with large, complex genomes. We discuss major considerations for developing and evaluating gLMs.

Thanks to a partnership with CNN Chile, our analyses were aired every Monday as part of a weekly program devoted to Plataforma Telar, with more details posted on our website and social media accounts. Our results were regularly met with high engagement, shared by media companies and personalities, and even by convention members.a Plataforma Telar thus had a noticeable impact on how people understood the convention (for an analysis of how data-driven political communication impacts public opinion, see Daud2). Given the diversity, scale and dynamics of the data, our cloud infrastructure was increasingly becoming unwieldy, with relevant information about particular entities (for example, convention members) scattered around different tables. In order to better structure these data, we structured these data as a knowledge graph, called TelarKG, which could then be queried using MillenniumDB: an open-source graph database also developed within the IMFD.

Graph databases (GDBs)13,30 have gained momentum with the rise of large unstructured repositories of information that emphasize relations between entities. Dozens of GDB management systems,8,22,25,31 prototypes,1,2,15,21 models and languages,3,10,12,14 large knowledge graphs like Wikidata,33 and efforts from companies like Apache, Facebook, Google, Microsoft, Neo4j, and Oracle, illustrate the growing interest in this technology. While the expressive power and flexibility of their data model and query languages is the key to their success, the efficiency challenges posed by their implementation is the main obstacle to the wider adoption of GDBs. Latin America has a long-standing tradition in fundamental research areas like database theory, string processing, information retrieval, and the design and analysis of algorithms and data structures--all of which are relevant for the development of GDBs. In the last few years, several researchers in Chile started collaborating on algorithms and systems for evaluating complex queries on large-scale GDBs.

fmeval is an open source library to evaluate large language models (LLMs) in a range of tasks. It helps practitioners evaluate their model for task performance and along multiple responsible AI dimensions. This paper presents the library and exposes its underlying design principles: simplicity, coverage, extensibility and performance. We then present how these were implemented in the scientific and engineering choices taken when developing fmeval. A case study demonstrates a typical use case for the library: picking a suitable model for a question answering task. We close by discussing limitations and further work in the development of the library. fmeval can be found at https://github.com/aws/fmeval.

This paper considers constrained spacecraft rendezvous and docking (RVD) in the setting of the Bicircular Restricted Four-Body Problem (BCR4BP), while accounting for attitude dynamics. We consider Line of Sight (LoS) cone constraints, thrust limits, thrust direction limits, and approach velocity constraints during RVD missions in a near rectilinear halo orbit (NRHO) in the Sun-Earth-Moon system. To enforce the constraints, the Time Shift Governor (TSG), which uses a time-shifted Chief spacecraft trajectory as a target reference for the Deputy spacecraft, is employed. The time shift is gradually reduced to zero so that the virtual target gradually evolves towards the Chief spacecraft as time goes by, and the RVD mission objective can be achieved. Numerical simulation results are reported to validate the proposed control method.

Most currently deployed large language models (LLMs) undergo continuous training or additional finetuning. By contrast, most research into LLMs' internal mechanisms focuses on models at one snapshot in time (the end of pre-training), raising the question of whether their results generalize to real-world settings. Existing studies of mechanisms over time focus on encoder-only or toy models, which differ significantly from most deployed models. In this study, we track how model mechanisms, operationalized as circuits, emerge and evolve across 300 billion tokens of training in decoder-only LLMs, in models ranging from 70 million to 2.8 billion parameters. We find that task abilities and the functional components that support them emerge consistently at similar token counts across scale. Moreover, although such components may be implemented by different attention heads over time, the overarching algorithm that they implement remains. Surprisingly, both these algorithms and the types of components involved therein can replicate across model scale. These results suggest that circuit analyses conducted on small models at the end of pre-training can provide insights that still apply after additional pre-training and over model scale.

With the increasing demand for multiple applications on internet of vehicles. It requires vehicles to carry out multiple computing tasks in real time. However, due to the insufficient computing capability of vehicles themselves, offloading tasks to vehicular edge computing (VEC) servers and allocating computing resources to tasks becomes a challenge. In this paper, a multi task digital twin (DT) VEC network is established. By using DT to develop offloading strategies and resource allocation strategies for multiple tasks of each vehicle in a single slot, an optimization problem is constructed. To solve it, we propose a multi-agent reinforcement learning method on the task offloading and resource allocation. Numerous experiments demonstrate that our method is effective compared to other benchmark algorithms.

Constrained Reinforcement Learning (CRL) tackles sequential decision-making problems where agents are required to achieve goals by maximizing the expected return while meeting domain-specific constraints, which are often formulated as expected costs. In this setting, policy-based methods are widely used since they come with several advantages when dealing with continuous-control problems. These methods search in the policy space with an action-based or parameter-based exploration strategy, depending on whether they learn directly the parameters of a stochastic policy or those of a stochastic hyperpolicy. In this paper, we propose a general framework for addressing CRL problems via gradient-based primal-dual algorithms, relying on an alternate ascent/descent scheme with dual-variable regularization. We introduce an exploration-agnostic algorithm, called C-PG, which exhibits global last-iterate convergence guarantees under (weak) gradient domination assumptions, improving and generalizing existing results. Then, we design C-PGAE and C-PGPE, the action-based and the parameter-based versions of C-PG, respectively, and we illustrate how they naturally extend to constraints defined in terms of risk measures over the costs, as it is often requested in safety-critical scenarios. Finally, we numerically validate our algorithms on constrained control problems, and compare them with state-of-the-art baselines, demonstrating their effectiveness.

This work presents a perception system applied to robotic manipulation, that is able to assist in navigation, household waste classification and collection in outdoor environments. This system is made up of optical tactile sensors, RGBD cameras and a LiDAR. These sensors are integrated on a mobile platform with a robot manipulator and a robotic gripper. Our system is divided in three software modules, two of them are vision-based and the last one is tactile-based. The vision-based modules use CNNs to localize and recognize solid household waste, together with the grasping points estimation. The tactile-based module, which also uses CNNs and image processing, adjusts the gripper opening to control the grasping from touch data. Our proposal achieves localization errors around 6 %, a recognition accuracy of 98% and ensures the grasping stability the 91% of the attempts. The sum of runtimes of the three modules is less than 750 ms.

Graph data structures offer a versatile and powerful means to model relationships and interconnections in various domains, promising substantial advantages in data representation, analysis, and visualization. In games, graph-based data structures are omnipresent and represent, for example, game economies, skill trees or complex, branching quest lines. With this paper, we propose G-PCGRL, a novel and controllable method for the procedural generation of graph data using reinforcement learning. Therefore, we frame this problem as manipulating a graph's adjacency matrix to fulfill a given set of constraints. Our method adapts and extends the Procedural Content Generation via Reinforcement Learning (PCGRL) framework and introduces new representations to frame the problem of graph data generation as a Markov decision process. We compare the performance of our method with the original PCGRL, the run time with a random search and evolutionary algorithm, and evaluate G-PCGRL on two graph data domains in games: game economies and skill trees. The results show that our method is capable of generating graph-based content quickly and reliably to support and inspire designers in the game creation process. In addition, trained models are controllable in terms of the type and number of nodes to be generated.