Artificial Intelligence (cs.AI)

Abstract: We present a temporally layered architecture (TLA) for temporally adaptive control with minimal energy expenditure. The TLA layers a fast and a slow policy together to achieve temporal abstraction that allows each layer to focus on a different time scale. Our design draws on the energy-saving mechanism of the human brain, which executes actions at different timescales depending on the environment's demands. We demonstrate that beyond energy saving, TLA provides many additional advantages, including persistent exploration, fewer required decisions, reduced jerk, and increased action repetition. We evaluate our method on a suite of continuous control tasks and demonstrate the significant advantages of TLA over existing methods when measured over multiple important metrics. We also introduce a multi-objective score to qualitatively assess continuous control policies and demonstrate a significantly better score for TLA. Our training algorithm uses minimal communication between the slow and fast layers to train both policies simultaneously, making it viable for future applications in distributed control.

Abstract: Accurately predicting the real-life performance of algorithms solving the Dial-a-Ride Problem (DARP) in the context of Mobility on Demand (MoD) systems with ridesharing requires evaluating them on representative instances. However, the benchmarking of state-of-the-art DARP solution methods has been limited to small, artificial instances or outdated non-public instances, hindering direct comparisons. With the rise of large MoD systems and the availability of open travel demand datasets for many US cities, there is now an opportunity to evaluate these algorithms on standardized, realistic, and representative instances. Despite the significant challenges involved in processing obfuscated and diverse datasets, we have developed a methodology using which we have created a comprehensive set of large-scale demand instances based on real-world data. These instances cover diverse use cases, one of which is demonstrated in an evaluation of two established DARP methods: the insertion heuristic and optimal vehicle-group assignment method. We publish the full results of both methods in a standardized format. The results show significant differences between areas in all measured quantities, emphasizing the importance of evaluating methods across different cities.

Abstract: This paper introduces the Fusemate probabilistic logic programming system. Fusemate's inference engine comprises a grounding component and a variable elimination method for probabilistic inference. Fusemate differs from most other systems by grounding the program in a bottom-up way instead of the common top-down way. While bottom-up grounding is attractive for a number of reasons, e.g., for dynamically creating distributions of varying support sizes, it makes it harder to control the amount of ground clauses generated. We address this problem by interleaving grounding (along program stratification) with a query-guided relevance test. This test prunes ground rules whose heads are inconsistent with the query dynamically extended by the ground rules so far. We present our method in detail and demonstrate it with examples that involve ``time'', such as (hidden) Markov models. Our experiments demonstrate competitive or better performance compared to a state-of-the probabilistic logic programming system, in particular for high branching problems.

Abstract: Aiding humans with scientific designs is one of the most exciting of artificial intelligence (AI) and machine learning (ML), due to their potential for the discovery of new drugs, design of new materials and chemical compounds, etc. However, scientific design typically requires complex domain knowledge that is not familiar to AI researchers. Further, scientific studies involve professional skills to perform experiments and evaluations. These obstacles prevent AI researchers from developing specialized methods for scientific designs. To take a step towards easy-to-understand and reproducible research of scientific design, we propose a benchmark for the inverse design of nanophotonic devices, which can be verified computationally and accurately. Specifically, we implemented three different nanophotonic design problems, namely a radiative cooler, a selective emitter for thermophotovoltaics, and structural color filters, all of which are different in design parameter spaces, complexity, and design targets. The benchmark environments are implemented with an open-source simulator. We further implemented 10 different inverse design algorithms and compared them in a reproducible and fair framework. The results revealed the strengths and weaknesses of existing methods, which shed light on several future directions for developing more efficient inverse design algorithms. Our benchmark can also serve as the starting point for more challenging scientific design problems. The code of IDToolkit is available at https://github.com/ThyrixYang/IDToolkit.

Abstract: The optimal multicast tree problem in the Software-Defined Networking (SDN) multicast routing is an NP-hard combinatorial optimization problem. Although existing SDN intelligent solution methods, which are based on deep reinforcement learning, can dynamically adapt to complex network link state changes, these methods are plagued by problems such as redundant branches, large action space, and slow agent convergence. In this paper, an SDN intelligent multicast routing algorithm based on deep hierarchical reinforcement learning is proposed to circumvent the aforementioned problems. First, the multicast tree construction problem is decomposed into two sub-problems: the fork node selection problem and the construction of the optimal path from the fork node to the destination node. Second, based on the information characteristics of SDN global network perception, the multicast tree state matrix, link bandwidth matrix, link delay matrix, link packet loss rate matrix, and sub-goal matrix are designed as the state space of intrinsic and meta controllers. Then, in order to mitigate the excessive action space, our approach constructs different action spaces at the upper and lower levels. The meta-controller generates an action space using network nodes to select the fork node, and the intrinsic controller uses the adjacent edges of the current node as its action space, thus implementing four different action selection strategies in the construction of the multicast tree. To facilitate the intelligent agent in constructing the optimal multicast tree with greater speed, we developed alternative reward strategies that distinguish between single-step node actions and multi-step actions towards multiple destination nodes.

Abstract: Strategic reasoning enables agents to cooperate, communicate, and compete with other agents in diverse situations. Existing approaches to solving strategic games rely on extensive training, yielding strategies that do not generalize to new scenarios or games without retraining. Large Language Models (LLMs), with their ability to comprehend and generate complex, context-rich language, could prove powerful as tools for strategic gameplay. This paper introduces an approach that uses pretrained LLMs with few-shot chain-of-thought examples to enable strategic reasoning for AI agents. Our approach uses systematically generated demonstrations of reasoning about states, values, and beliefs to prompt the model. Using extensive variations of simple matrix games, we show that strategies that are derived based on systematically generated prompts generalize almost perfectly to new game structures, alternate objectives, and hidden information. Additionally, we demonstrate our approach can lead to human-like negotiation strategies in realistic scenarios without any extra training or fine-tuning. Our results highlight the ability of LLMs, guided by systematic reasoning demonstrations, to adapt and excel in diverse strategic scenarios.

Abstract: The rapid advancement of artificial intelligence (AI) systems suggests that artificial general intelligence (AGI) systems may soon arrive. Many researchers are concerned that AIs and AGIs will harm humans via intentional misuse (AI-misuse) or through accidents (AI-accidents). In respect of AI-accidents, there is an increasing effort focused on developing algorithms and paradigms that ensure AI systems are aligned to what humans intend, e.g. AI systems that yield actions or recommendations that humans might judge as consistent with their intentions and goals. Here we argue that alignment to human intent is insufficient for safe AI systems and that preservation of long-term agency of humans may be a more robust standard, and one that needs to be separated explicitly and a priori during optimization. We argue that AI systems can reshape human intention and discuss the lack of biological and psychological mechanisms that protect humans from loss of agency. We provide the first formal definition of agency-preserving AI-human interactions which focuses on forward-looking agency evaluations and argue that AI systems - not humans - must be increasingly tasked with making these evaluations. We show how agency loss can occur in simple environments containing embedded agents that use temporal-difference learning to make action recommendations. Finally, we propose a new area of research called "agency foundations" and pose four initial topics designed to improve our understanding of agency in AI-human interactions: benevolent game theory, algorithmic foundations of human rights, mechanistic interpretability of agency representation in neural-networks and reinforcement learning from internal states.