This work identifies 18 foundational challenges in assuring the alignment and safety of large language models (LLMs). These challenges are organized into three different categories: scientific understanding of LLMs, development and deployment methods, and sociotechnical challenges. Based on the identified challenges, we pose $200+$ concrete research questions.

Stellar collisions in dense galactic nuclei might play an important role in fueling supermassive black holes (SMBHs) and shaping their environments. The gas released during these collisions can contribute to SMBH accretion, influencing phenomena such as active galactic nuclei and tidal disruption events of the remnants. We address the challenge of rapidly and accurately predicting the outcomes of stellar collisionsincluding remnant masses and unbound gasacross a broad parameter space of initial conditions. Existing smoothed-particle-hydrodynamic (SPH) simulation techniques, while detailed, are too resource-intensive for exploratory studies or real-time applications. We develop a machine learning framework trained on a dataset of $\sim 16,000$ SPH simulations of main-sequence star collisions. By extracting physically meaningful parameters (e.g., masses, radii, impact parameters, and virial ratios) and employing gradient-boosted regression trees with Huber loss, we create a model that balances accuracy and computational efficiency. The method includes logarithmic transforms to handle dynamic ranges and regularization to ensure physical plausibility. The model achieves predictions of collision outcomes (remnant masses, and unbound mass) with very low mean absolute errors respect to the typical mass scale. It operates in fractions of a second, enabling large-scale parameter studies and real-time applications. Parameter importance analysis reveals that the impact parameter and the relative velocity dominate outcomes, aligning with theoretical expectations. Our approach provides a scalable tool for studying stellar collisions in galactic nuclei. The rapid predictions facilitate investigations into gas supply for SMBH accretion and the cumulative effects of collisions over cosmic time, particularly relevant to address the growth of SMBHs.

The increasing prevalence of mental health disorders globally highlights the urgent need for effective digital screening methods that can be used in multilingual contexts. Most existing studies, however, focus on English data, overlooking critical mental health signals that may be present in non-English texts. To address this important gap, we present the first survey on the detection of mental health disorders using multilingual social media data. We investigate the cultural nuances that influence online language patterns and self-disclosure behaviors, and how these factors can impact the performance of NLP tools. Additionally, we provide a comprehensive list of multilingual data collections that can be used for developing NLP models for mental health screening. Our findings can inform the design of effective multilingual mental health screening tools that can meet the needs of diverse populations, ultimately improving mental health outcomes on a global scale.

Data science aims to extract insights from data to support decision-making processes. Recently, Large Language Models (LLMs) have been increasingly used as assistants for data science, by suggesting ideas, techniques and small code snippets, or for the interpretation of results and reporting. Proper automation of some data-science activities is now promised by the rise of LLM agents, i.e., AI systems powered by an LLM equipped with additional affordances--such as code execution and knowledge bases--that can perform self-directed actions and interact with digital environments. In this paper, we survey the evaluation of LLM assistants and agents for data science. We find (1) a dominant focus on a small subset of goal-oriented activities, largely ignoring data management and exploratory activities; (2) a concentration on pure assistance or fully autonomous agents, without considering intermediate levels of human-AI collaboration; and (3) an emphasis on human substitution, therefore neglecting the possibility of higher levels of automation thanks to task transformation.

In a seminal paper [8] it was shown that Heisenberg-limited measurements could be achieved without using entangled states by coupling the quantum resources to a common environment that could be measured, at least, in part. The authors also claimed that their method would be robust under decoherence, and, in particular, applied it to derive an analytical expression to measure the change in length of an optical cavity within the Tavis-Cummings model using the superradiance approximation, which would still have a 1/N scaling. Here, we show that the analytical derivations in [8] are incompatible with the cavity superradiance, leading to the standard quantum-limited scaling 1/\sqrt{N}.

Aligning large language models (LLMs) with human values is essential for their safe deployment and widespread adoption. Current LLM safety benchmarks often focus solely on the refusal of individual problematic queries, which overlooks the importance of the context where the query occurs and may cause undesired refusal of queries under safe contexts that diminish user experience. Addressing this gap, we introduce CASE-Bench, a Context-Aware SafEty Benchmark that integrates context into safety assessments of LLMs. CASE-Bench assigns distinct, formally described contexts to categorized queries based on Contextual Integrity theory. Additionally, in contrast to previous studies which mainly rely on majority voting from just a few annotators, we recruited a sufficient number of annotators necessary to ensure the detection of statistically significant differences among the experimental conditions based on power analysis. Our extensive analysis using CASE-Bench on various open-source and commercial LLMs reveals a substantial and significant influence of context on human judgments (p<0.0001 from a z-test), underscoring the necessity of context in safety evaluations. We also identify notable mismatches between human judgments and LLM responses, particularly in commercial models within safe contexts.

Today, Wi-Fi is over 25 years old. Yet, despite sharing the same branding name, today's Wi-Fi boasts entirely new capabilities that were not even on the roadmap 25 years ago. This article aims to provide a holistic and comprehensive technical and historical tutorial on Wi-Fi, beginning with IEEE 802.11b (Wi-Fi 1) and looking forward to IEEE 802.11bn (Wi-Fi 8). This is the first tutorial article to span these eight generations. Rather than a generation-by-generation exposition, we describe the key mechanisms that have advanced Wi-Fi. We begin by discussing spectrum allocation and coexistence, and detailing the IEEE 802.11 standardization cycle. Second, we provide an overview of the physical layer and describe key elements that have enabled data rates to increase by over 1,000x. Third, we describe how Wi-Fi Medium Access Control has been enhanced from the original Distributed Coordination Function to now include capabilities spanning from frame aggregation to wideband spectrum access. Fourth, we describe how Wi-Fi 5 first broke the one-user-at-a-time paradigm and introduced multi-user access. Fifth, given the increasing use of mobile, battery-powered devices, we describe Wi-Fi's energy-saving mechanisms over the generations. Sixth, we discuss how Wi-Fi was enhanced to seamlessly aggregate spectrum across 2.4 GHz, 5 GHz, and 6 GHz bands to improve throughput, reliability, and latency. Finally, we describe how Wi-Fi enables nearby Access Points to coordinate in order to improve performance and efficiency. In the Appendix, we further discuss Wi-Fi developments beyond 802.11bn, including integrated mmWave operations, sensing, security and privacy extensions, and the adoption of AI/ML.

Initialization plays a critical role in Deep Neural Network training, directly influencing convergence, stability, and generalization. Common approaches such as Glorot and He initializations rely on randomness, which can produce uneven weight distributions across layer connections. In this paper, we introduce the Sinusoidal initialization, a novel deterministic method that employs sinusoidal functions to construct structured weight matrices expressly to improve the spread and balance of weights throughout the network while simultaneously fostering a more uniform, well-conditioned distribution of neuron activation states from the very first forward pass. Because Sinusoidal initialization begins with weights and activations that are already evenly and efficiently utilized, it delivers consistently faster convergence, greater training stability, and higher final accuracy across a wide range of models, including convolutional neural networks, vision transformers, and large language models. On average, our experiments show an increase of 4.9% in final validation accuracy and 20.9% in convergence speed. By replacing randomness with structure, this initialization provides a stronger and more reliable foundation for Deep Learning systems.

Adaptation of foundation models using low-rank methods is a widespread approach. Another way to adapt these models is to employ orthogonal fine-tuning methods, which are less time and memory efficient despite their good generalization properties. In this work, we propose Householder Orthogonal Fine-tuning (HOFT), a novel orthogonal fine-tuning method that aims to alleviate time and space complexity. Moreover, some theoretical properties of the orthogonal fine-tuning paradigm are explored. From this exploration, Scaled Householder Orthogonal Fine-tuning (SHOFT) is proposed. Both HOFT and SHOFT are evaluated in downstream tasks, namely commonsense reasoning, machine translation, subject-driven generation and mathematical reasoning. Compared with state-of-the-art adaptation methods, HOFT and SHOFT show comparable or better results.

Modular quantum computing architectures offer a promising alternative to monolithic designs for overcoming the scaling limitations of current quantum computers. To achieve scalability beyond small prototypes, quantum architectures are expected to adopt a modular approach, featuring clusters of tightly connected quantum bits with sparser connections between these clusters. Efficiently distributing qubits across multiple processing cores is critical for improving quantum computing systems' performance and scalability. To address this challenge, we propose the Hungarian Qubit Assignment (HQA) algorithm, which leverages the Hungarian algorithm to improve qubit-to-core assignment. The HQA algorithm considers the interactions between qubits over the entire circuit, enabling fine-grained partitioning and enhanced qubit utilization. We compare the HQA algorithm with state-of-the-art alternatives through comprehensive experiments using both real-world quantum algorithms and random quantum circuits. The results demonstrate the superiority of our proposed approach, outperforming existing methods, with an average improvement of 1.28$\times$.

Context: The detection of the highest energy neutrino observed to date by KM3NeT, with an estimated energy of 220 PeV, opens up new possibilities for the study and identification of the astrophysical sources responsible for a diffuse flux of such ultra-high-energy neutrinos, among which gamma-ray bursts are longstanding candidates. Aims: Based on the event KM3-230213A, we derive constraints on the baryon loading and density of the surrounding environment in models of blastwaves in long-duration gamma-ray bursts. Methods: We compute the diffuse flux from gamma-ray burst blastwaves, either expanding in a constant density interstellar medium or developing in a radially decreasing density of a wind-like environment surrounding the gamma-ray burst progenitor star, by taking into account the expected neutrino spectra and luminosity function. We use a Poisson likelihood method to constrain the blastwave model parameters by calculating the expected number of neutrino events within the 90% confidence level energy range of KM3-230213A and by using the joint exposure of KM3NeT/ARCA, IceCube and Pierre Auger. Results: We constrain the baryon loading to be $\leq \{392, 131, 39, 13\}$ at 90% confidence level, which is inversely proportional to a varying interstellar medium particle density of $\{1, 3, 10, 30\}$ cm$^{-3}$. In the wind-like environment case, the baryon loading is $\leq \{20, 50, 100\}$ at 90% confidence level, which is proportional to the sixth power of a varying density parameter of $\{0.05, 0.06, 0.07\}$.

The gravitational capture of a stellar-mass compact object (CO) by a supermassive black hole is a unique probe of gravity in the strong field regime. Because of the large mass ratio, we call these sources extreme-mass ratio inspirals (EMRIs). In a similar manner, COs can be captured by intermediate-mass black holes in globular clusters or dwarf galaxies. The mass ratio in this case is lower, and hence we refer to the system as an intermediate-mass ratio inspiral (IMRI). Also, sub-stellar objects such as a brown dwarf, with masses much lighter than our Sun, can inspiral into supermassive black holes such as Sgr A* at our Galactic centre. In this case, the mass ratio is extremely large and, hence, we call this system ab extremely-large mass ratio inspirals (XMRIs). All of these sources of gravitational waves will provide us with a collection of snapshots of spacetime around a supermassive black hole that will allow us to do a direct mapping of warped spacetime around the supermassive black hole, a live cartography of gravity in this extreme gravity regime. E/I/XMRIs will be detected by the future space-borne observatories like LISA. There has not been any other probe conceived, planned or even thought of ever that can do the science that we can do with these inspirals. We will discuss them from a viewpoint of relativistic astrophysics.

Multi-agent systems (MAS) have gained relevance in the field of artificial intelligence by offering tools for modelling complex environments where autonomous agents interact to achieve common or individual goals. In these systems, norms emerge as a fundamental component to regulate the behaviour of agents, promoting cooperation, coordination and conflict resolution. This article presents a systematic review, following the PRISMA method, on the emergence of norms in MAS, exploring the main mechanisms and factors that influence this process. Sociological, structural, emotional and cognitive aspects that facilitate the creation, propagation and reinforcement of norms are addressed. The findings highlight the crucial role of social network topology, as well as the importance of emotions and shared values in the adoption and maintenance of norms. Furthermore, opportunities are identified for future research that more explicitly integrates emotional and ethical dynamics in the design of adaptive normative systems. This work provides a comprehensive overview of the current state of research on norm emergence in MAS, serving as a basis for advancing the development of more efficient and flexible systems in artificial and real-world contexts.

Many computationally hard problems can be encoded in quantum Hamiltonians. The solution to these problems is given by the ground states of these Hamiltonians. A state-of-the-art algorithm for finding the ground state of a Hamiltonian is the so-called Quantum Imaginary Time Evolution (QITE) which approximates imaginary time evolution by a unitary evolution that can be implemented in quantum hardware. In this paper, we review the original algorithm together with a comprehensive computer program, as well as, the variational version of it.