Kazantzis-Kravaris/Luenberger (KKL) observers are a class of state observers for nonlinear systems that rely on an injective map to transform the nonlinear dynamics into a stable quasi-linear latent space, from where the state estimate is obtained in the original coordinates via a left inverse of the transformation map. Current learning-based methods for these maps are designed exclusively for autonomous systems and do not generalize well to controlled or non-autonomous systems. In this paper, we propose two learning-based designs […]

Ver mais

Every vector search implementation seems to follow the exact same trajectory: Teams do a proof of concept, it works great on their test setup, they ship it, and 3 months later the cluster is already having performance issues and we are already into our first round of “architectural improvements”. Production scale completely blew off all the calculations. Here’s what nobody realizes: scaling vector search isn’t one problem. It’s three. You hit a compute wall first. Solve that, and […]

Ver mais

arXiv:2603.10048v1 Announce Type: new Abstract: Sharpness-Aware Minimization (SAM) enhances generalization by minimizing the maximum training loss within a predefined neighborhood around the parameters. However, its practical implementation approximates this as gradient ascent(s) followed by applying the gradient at the ascent point to update the current parameters. This practice can be justified as approximately optimizing the objective by neglecting the (full) derivative of the ascent point with respect to the current parameters. Nevertheless, a direct and intuitive understanding of […]

Ver mais

In this quiz, you’ll test your understanding of Python tuples. By working through this quiz, you’ll revisit various ways to interact with Python tuples. You’ll also practice recognizing common features and gotchas. [ Improve Your Python With 🐍 Python Tricks 💌 – Get a short & sweet Python Trick delivered to your inbox every couple of days. >> Click here to learn more and see examples ]

Ver mais

Continual instruction tuning (CIT) requires multimodal large language models (MLLMs) to adapt to a stream of tasks without forgetting prior capabilities. A common strategy is to isolate updates by routing inputs to different LoRA experts. However, existing LoRA-based Mixture-of-Experts (MoE) methods often jointly update the router and experts in an indiscriminate way, causing the router’s preferences to co-drift with experts’ adaptation pathways and gradually deviate from early-stage input-expert specialization. We term this phenomenon Misaligned Co-drift, which blurs expert […]

Ver mais

arXiv:2505.19134v2 Announce Type: replace-cross Abstract: Human-annotated data plays a vital role in training large language models (LLMs), such as supervised fine-tuning and human preference alignment. However, it is not guaranteed that paid human annotators produce high-quality data. In this paper, we study how to incentivize human annotators to do so. We start from a principal-agent model to model the dynamics between the company (the principal) and the annotator (the agent), where the principal can only monitor the annotation […]

Ver mais

arXiv:2601.08887v1 Announce Type: new Abstract: Efficient load-balancing mechanisms are critical for maximizing performance and increasing the quality of service (QoS) of data center networks (DCNs). Obtaining the optimal QoS while minimizing resource consumption remains a significant challenge. This paper proposes the streamlined pathway (SP) model, which is a flow scheduling solution that requires minimal statistical knowledge of the DCN data plane. The SP model utilizes the software-defined networks (SDN) paradigm with less information gathered from the DCN data […]

Ver mais

arXiv:2602.13262v1 Announce Type: new Abstract: Frontier language models improve with additional test-time computation, but serial reasoning or uncoordinated parallel sampling can be compute-inefficient under fixed inference budgets. We propose SELFCEST, which equips a base model with the ability to spawn same-weight clones in separate parallel contexts by agentic reinforcement learning. Training is end-to-end under a global task reward with shared-parameter rollouts, yielding a learned controller that allocates both generation and context budget across branches. Across challenging math reasoning […]

Ver mais

arXiv:2602.00918v1 Announce Type: new Abstract: Early Classification of Time Series (ECTS) addresses decision-making problems in which predictions must be made as early as possible while maintaining high accuracy. Most existing ECTS methods assume that the time-dependent decision costs governing the learning objective are known, fixed, and correctly specified. In practice, however, these costs are often uncertain and may change over time, leading to mismatches between training-time and deployment-time objectives. In this paper, we study ECTS under two practically […]

Ver mais