Long-horizon tasks that require sustained reasoning and multiple tool interactions remain challenging for LLM agents: small errors compound across steps, and even state-of-the-art models often hallucinate or lose coherence.We identify context management as the central bottleneck-extended histories cause agents to overlook critical evidence or become distracted by irrelevant information, thus failing to replan or reflect from previous mistakes.To address this, we propose COMPASS (Context-Organized Multi-Agent Planning and Strategy System), a lightweight hierarchical framework that separates tactical execution, strategic oversight, and context organization into three specialized components: (1) a Main Agent that performs reasoning and tool use, (2) a Meta-Thinker that monitors progress and issues strategic interventions, and (3) a Context Manager that maintains concise, relevant progress briefs for different reasoning stages.Across three challenging benchmarks-GAIA, BrowseComp, and Humanity's Last Exam-COMPASS improves accuracy by up to 20% relative to both single-and multi-agent baselines.We further introduce a test-time scaling extension that elevates performance to match established DeepResearch agents, and a posttraining pipeline that delegates context management to smaller models for enhanced efficiency.

This study addresses the challenge of multi-robot cooperative exploration under limited local observations in environments with dynamic robot populations. To achieve efficient area coverage within constrained timeframes, we propose the Multi-Robot Informative Planner (MIP), a novel reinforcement learning (RL)-based planning module. The core component of MIP is the Neighborhood Information Aggregator, which employs a graph neural network (GNN) to integrate local neighborhood information for each robot. Our design enhances sample efficiency by minimizing information requirements while ensuring scalability across environments with varying robot numbers. To generate high-quality, expressive neighborhood feature representations, we utilize Graphical Mutual Information (GMI) to maximize the correlation between neighboring robots’ input features and their high-level hidden representations. Furthermore, MIP incorporates the Spatial-Neighborhood Transformer, which captures spatial features and inter-robot interactions through spatial self-attention mechanisms. These components collectively form the Multi-Robot Neural Informative Mapping (MRNIM) framework, outperforming traditional benchmarks in Habitat simulator.

The mainstream paradigm behind continual learning has been to adapt the model parameters to non-stationary data distributions, where catastrophic forgetting is the central challenge. Typical methods rely on a rehearsal buffer or known task identity at test time to retrieve learned knowl-edge and address forgetting, while this work presents a new paradigm for continual learning that aims to train a more succinct memory system without accessing task identity at test time. Our method learns to dynamically prompt (L2P) a pre-trained model to learn tasks sequen-tially under different task transitions. In our proposed framework, prompts are small learnable parameters, which are maintained in a memory space. The objective is to optimize prompts to instruct the model prediction and ex-plicitly manage task-invariant and task-specific knowledge while maintaining model plasticity. We conduct comprehen-sive experiments under popular image classification bench-marks with different challenging continual learning set-tings, where L2P consistently outperforms prior state-of-the-art methods. Surprisingly, L2P achieves competitive results against rehearsal-based methods even without a re-hearsal buffer and is directly applicable to challenging task-agnostic continual learning. Source code is available at https://github.com/google-research/12p.

Despite the common practice of oversubscription, power capacity is largely under-utilized in data centers. A significant factor driving this under-utilization is fluctuation of the aggregate power demand, resulting in unused “spot (power) capacity”. In this paper, we tap into spot capacity for improving power infrastructure utilization in multi-tenant data centers, an important but under-explored type of data center where multiple tenants house their own physical servers. We propose a novel market, called SpotDC, to allocate spot capacity to tenants on demand. Specifically, SpotDC extracts tenants' racklevel spot capacity demand through an elastic demand function, based on which the operator sets the market price for spot capacity allocation. We evaluate SpotDC using both testbed experiments and simulations, demonstrating that SpotDC improves power infrastructure utilization and creates a “win-win” situation: the data center operator increases its profit (by nearly 10%), while tenants improve their performance (by 1.2-1.8x on average compared to the no spot capacity case, yet at a marginal cost).

Power oversubscription in data centers may occasionally trigger an emergency when the aggregate power demand exceeds the capacity. Handling such an emergency requires a graceful power capping solution that minimizes the performance loss. In this paper, we study power capping in a multi-tenant data center where the operator supplies power to multiple tenants that manage their own servers. Unlike owner-operated data centers, the operator lacks control over tenants' servers. To address this challenge, we propose a novel market mechanism based on supply function bidding, called COOP, to financially incentivize and coordinate tenants' power reduction for minimizing total performance loss (quantified in performance cost) while satisfying multiple power capping constraints. We build a prototype to show that COOP is efficient in terms of minimizing the total performance cost, even compared to the ideal but infeasible case that assumes the operator has full control over tenants' servers. We also demonstrate that COOP is "win-win", increasing the operator's profit (through oversubscription) and reducing tenants' cost (through financial compensation for their power reduction during emergencies).

This paper studies design challenges faced by a geo-distributed cloud data market: which data to purchase (data purchasing) and where to place/replicate the data (data placement). We show that the joint problem of data purchasing and data placement within a cloud data market is NP-hard in general. However, we give a provably optimal algorithm for the case of a data market made up of a single data center, and then generalize the structure from the single data center setting and propose Datum, a near-optimal, polynomial-time algorithm for a geo-distributed data market.

As clusters continue to grow in size and complexity, providing scalable and predictable performance is an increasingly important challenge. A crucial roadblock to achieving predictable performance is stragglers, i.e., tasks that take significantly longer than expected to run. At this point, speculative execution has been widely adopted to mitigate the impact of stragglers. However, speculation mechanisms are designed and operated independently of job scheduling when, in fact, scheduling a speculative copy of a task has a direct impact on the resources available for other jobs. In this work, we present Hopper, a job scheduler that is speculation-aware, i.e., that integrates the tradeoffs associated with speculation into job scheduling decisions. We implement both centralized and decentralized prototypes of the Hopper scheduler and show that 50% (66%) improvements over state-of-the-art centralized (decentralized) schedulers and speculation strategies can be achieved through the coordination of scheduling and speculation.

Accurate performance evaluation for network algorithms is vital to meet various requirements of different applications, such as QoS, network security, traffic engineering. Although worst-case and average-case analysis are widely used in algorithm evaluation, they are often insufficient due to the lack of practicality. Smoothed Analysis (SA) introduces a new concept of smoothed complexity, remedying the shortcomings in worst-case and average-case analysis. However, recent research towards SA focuses on theoretical evaluation, and thus those methods tend to be too complicated for the analysis of network algorithms. To address the problem, Sampling-based Smoothed Analysis (SSA) for network algorithm evaluation is proposed. SSA extends feasibility for practical performance evaluation and achieves promising experimental results. As examples, two algorithms for typical network problem are evaluated using the proposed SSA framework, and the results explicitly illustrate their significant performance difference in spite of the same theoretical worst-case complexity. Besides evaluation accuracy, SSA also provide more insight for algorithms to facilitate current algorithms improvement and new algorithms design.