Planning has been a cornerstone of artificial intelligence for solving complex problems, and recent progress in LLM-based multi-agent frameworks have begun to extend this capability. However, the role of human-like memory within these frameworks remains largely unexplored. Understanding how agents coordinate through memory is critical for natural language planning, where iterative reasoning, constraint tracking, and error correction drive the success. Inspired by working memory model in cognitive psychology, we present EvoMem, a multi-agent framework built on a dual-evolving memory mechanism. The framework consists of three agents (Constraint Extractor, Verifier, and Actor) and two memory modules: Constraint Memory (CMem), which evolves across queries by storing task-specific rules and constraints while remains fixed within a query, and Query-feedback Memory (QMem), which evolves within a query by accumulating feedback across iterations for solution refinement. Both memory modules are reset at the end of each query session. Evaluations on trip planning, meeting planning, and calendar scheduling show consistent performance improvements, highlighting the effectiveness of EvoMem. This success underscores the importance of memory in enhancing multi-agent planning.

Recent advancements in Large Language Models (LLMs) have created new opportunities to enhance performance on complex reasoning tasks by leveraging test-time computation. However, conventional approaches such as repeated sampling with majority voting or reward model scoring, often face diminishing returns as test-time compute scales, in addition to requiring costly task-specific reward model training. In this paper, we present Self-Enhanced Test-Time Scaling (SETS), a novel method that leverages the self-verification and self-correction capabilities of recent advanced LLMs to overcome these limitations. SETS integrates sampling, self-verification, and self-correction into a unified framework, enabling efficient and scalable test-time computation for improved capabilities at complex tasks. Through extensive experiments on challenging planning and reasoning benchmarks, compared to the alternatives, we demonstrate that SETS achieves significant performance improvements and more favorable test-time scaling laws.

The generative AI revolution is underway, with these bots now taking care of everything from coding apps to making movies (or at least attempting to). One way you'll sometimes see these AI chatbots used is as smart travel assistants, giving you recommendations on the locations to stay at, eat at, and tour around in just about any location you can name. There's no doubt that AI can be helpful here, in a variety of different ways, but it's also important to remember the limitations of the technology. These chatbots have never visited the places they're talking about--they don't know what fine dining is, or what a cozy hideaway is, they're just regurgitating text they've found on the web (albeit in a smart and natural way). By all means enlist the help of a generative AI bot when you're planning a trip, but be aware of the do's and don'ts.

We introduce NATURAL PLAN, a realistic planning benchmark in natural language containing 3 key tasks: Trip Planning, Meeting Planning, and Calendar Scheduling. We focus our evaluation on the planning capabilities of LLMs with full information on the task, by providing outputs from tools such as Google Flights, Google Maps, and Google Calendar as contexts to the models. This eliminates the need for a tool-use environment for evaluating LLMs on Planning. We observe that NATURAL PLAN is a challenging benchmark for state of the art models. For example, in Trip Planning, GPT-4 and Gemini 1.5 Pro could only achieve 31.1% and 34.8% solve rate respectively. We find that model performance drops drastically as the complexity of the problem increases: all models perform below 5% when there are 10 cities, highlighting a significant gap in planning in natural language for SoTA LLMs. We also conduct extensive ablation studies on NATURAL PLAN to further shed light on the (in)effectiveness of approaches such as self-correction, few-shot generalization, and in-context planning with long-contexts on improving LLM planning.

With recent advancements in the field of communications and the Internet of Things, vehicles are becoming more aware of their environment and are evolving towards full autonomy. Vehicular communication opens up the possibility for vehicle-to-infrastructure interaction, where vehicles could share information with components such as cameras, traffic lights, and signage that support a countrys road system. As a result, vehicles are becoming more than just a means of transportation; they are collecting, processing, and transmitting massive amounts of data used to make driving safer and more convenient. With 5G cellular networks and beyond, there is going to be more data bandwidth available on our roads, but it may be heterogeneous because of limitations like line of sight, infrastructure, and heterogeneous traffic on the road. This paper addresses the problem of route planning for autonomous vehicles in urban areas accounting for both driving time and data transfer needs. We propose a novel reinforcement learning solution that prioritizes high bandwidth roads to meet a vehicles data transfer requirement, while also minimizing driving time. We compare this approach to traffic-unaware and bandwidth-unaware baselines to show how much better it performs under heterogeneous traffic. This solution could be used as a starting point to understand what good policies look like, which could potentially yield faster, more efficient heuristics in the future.

Google has announced a slew of new search updates, ranging from travel planning to clothes shopping -- oh, and a bit of skin abnormality checking for good measure. That's right, Lens is no longer just for naming a plant or historical object but will now identify things about your skin. You simply upload a picture into Lens, and it will show you similar images. This update might be good for determining if you have a tick bite, but, like any Google searches when you're not feeling well, it could lead you down a pretty scary rabbit hole. Try to consult with a doctor if there are any spots you're unsure about across your skin.

Despite a tremendous amount of work in the literature and in the commercial sectors, current approaches to multi-modal trip planning still fail to consistently generate plans that users deem optimal in practice. We believe that this is due to the fact that current planners fail to capture the true preferences of users, e.g., their preferences depend on aspects that are not modeled. An example of this could be a preference not to walk through an unsafe area at night. We present a novel multi-modal trip planner that allows users to upload auxiliary geographic data (e.g., crime rates) and to specify temporal constraints and preferences over these data in combination with typical metrics such as time and cost. Concretely, our planner supports the modes walking, biking, driving, public transit, and taxi, uses linear temporal logic to capture temporal constraints, and preferential cost functions to represent preferences. We show by examples that this allows the expression of very interesting preferences and constraints that, naturally, lead to quite diverse optimal plans.

Despite a tremendous amount of work in the literature and in the commercial sectors, current approaches to multi-modal trip planning still fail to consistently generate plans that users deem optimal in practice. We believe that this is due to the fact that current planners fail to capture the true preferences of users, e.g., their preferences depend on aspects that are not modeled. An example of this could be a preference not to walk through an unsafe area at night. We present a novel multi-modal trip planner that allows users to up- load auxiliary geographic data (e.g., crime rates) and to specify temporal constraints and preferences over these data in combination with typical metrics such as time and cost. Concretely, our planner supports the modes walking, biking, driving, public transit, and taxi, uses linear temporal logic to capture temporal constraints, and preferential cost functions to represent preferences. We show by examples that this allows the expression of very interesting preferences and constraints that, naturally, lead to quite diverse optimal plans.

Whether for research, purchasing, education or in conjuncation with a family vacation, going to and attending robotics-related conferences and trade shows can provide real personal and business benefits, particularly when the venue is in a foreign location.

Robert Martin will show slides from his recent journey to Lebanon as well as a short visit to Jerusalem. Learn where and why avalanches occur, how to manage risk and simple ways to avoid avalanche hazards. Whitney expert and mountain guide Kurt Wedberg will discuss gear, trip planning and popular routes to the summit. Please email announcements at least three weeks before the event to travel@latimes.com.