Strategic decisions are often made over multiple periods of time, wherein decisions made earlier impact a competitor's success in later stages. In this paper, we study these dynamics in General Lotto games, a class of models describing the competitive allocation of resources between two opposing players. We propose a two-stage formulation where one of the players has reserved resources that can be strategically pre-allocated across the battlefields in the first stage of the game as reinforcements. The players then simultaneously allocate their remaining real-time resources, which can be randomized, in a decisive final stage. Our main contributions provide complete characterizations of the optimal reinforcement strategies and resulting equilibrium payoffs in these multi-stage General Lotto games. Interestingly, we determine that real-time resources are at least twice as effective as reinforcement resources when considering equilibrium payoffs.
相關內容
Notions of counterfactual invariance (CI) have proven essential for predictors that are fair, robust, and generalizable in the real world. We propose graphical criteria that yield a sufficient condition for a predictor to be counterfactually invariant in terms of a conditional independence in the observational distribution. In order to learn such predictors, we propose a model-agnostic framework, called Counterfactually Invariant Prediction (CIP), building on the Hilbert-Schmidt Conditional Independence Criterion (HSCIC), a kernel-based conditional dependence measure. Our experimental results demonstrate the effectiveness of CIP in enforcing counterfactual invariance across various simulated and real-world datasets including scalar and multi-variate settings.
A primary criticism towards language models (LMs) is their inscrutability. This paper presents evidence that, despite their size and complexity, LMs sometimes exploit a simple computational mechanism to solve one-to-one relational tasks (e.g., capital_of(Poland)=Warsaw). We investigate a range of language model sizes (from 124M parameters to 176B parameters) in an in-context learning setting, and find that for a variety of tasks (involving capital cities, upper-casing, and past-tensing) a key part of the mechanism reduces to a simple linear update typically applied by the feedforward (FFN) networks. These updates also tend to promote the output of the relation in a content-independent way (e.g., encoding Poland:Warsaw::China:Beijing), revealing a predictable pattern that these models take in solving these tasks. We further show that this mechanism is specific to tasks that require retrieval from pretraining memory, rather than retrieval from local context. Our results contribute to a growing body of work on the mechanistic interpretability of LLMs, and offer reason to be optimistic that, despite the massive and non-linear nature of the models, the strategies they ultimately use to solve tasks can sometimes reduce to familiar and even intuitive algorithms.
In many interactive decision-making settings, there is latent and unobserved information that remains fixed. Consider, for example, a dialogue system, where complete information about a user, such as the user's preferences, is not given. In such an environment, the latent information remains fixed throughout each episode, since the identity of the user does not change during an interaction. This type of environment can be modeled as a Latent Markov Decision Process (LMDP), a special instance of Partially Observed Markov Decision Processes (POMDPs). Previous work established exponential lower bounds in the number of latent contexts for the LMDP class. This puts forward a question: under which natural assumptions a near-optimal policy of an LMDP can be efficiently learned? In this work, we study the class of LMDPs with {\em prospective side information}, when an agent receives additional, weakly revealing, information on the latent context at the beginning of each episode. We show that, surprisingly, this problem is not captured by contemporary settings and algorithms designed for partially observed environments. We then establish that any sample efficient algorithm must suffer at least $\Omega(K^{2/3})$-regret, as opposed to standard $\Omega(\sqrt{K})$ lower bounds, and design an algorithm with a matching upper bound.
In the ever-evolving field of Artificial Intelligence, a critical challenge has been to decipher the decision-making processes within the so-called "black boxes" in deep learning. Over recent years, a plethora of methods have emerged, dedicated to explaining decisions across diverse tasks. Particularly in tasks like image classification, these methods typically identify and emphasize the pivotal pixels that most influence a classifier's prediction. Interestingly, this approach mirrors human behavior: when asked to explain our rationale for classifying an image, we often point to the most salient features or aspects. Capitalizing on this parallel, our research embarked on a user-centric study. We sought to objectively measure the interpretability of three leading explanation methods: (1) Prototypical Part Network, (2) Occlusion, and (3) Layer-wise Relevance Propagation. Intriguingly, our results highlight that while the regions spotlighted by these methods can vary widely, they all offer humans a nearly equivalent depth of understanding. This enables users to discern and categorize images efficiently, reinforcing the value of these methods in enhancing AI transparency.
Pilot contamination is a critical issue in distributed massive MIMO networks, where the reuse of pilot sequences due to limited availability of orthogonal pilots for channel estimation leads to performance degradation. In this work, we propose a novel distributed pilot assignment scheme to effectively mitigate the impact of pilot contamination. Our proposed scheme not only reduces signaling overhead, but it also enhances fault-tolerance. Extensive numerical simulations are conducted to evaluate the performance of the proposed scheme. Our results establish that the proposed scheme outperforms existing centralized and distributed schemes in terms of mitigating pilot contamination and significantly enhancing network throughput.
Factors within a large-scale software system that simultaneously interact and strongly impact the system's response under a configuration are often difficult to identify. Although screening such a system for the existence of such interactions is important, determining their location is more useful for system engineers. Combinatorial interaction testing (CIT) concerns creation of test suites that nonadaptively either detect or locate the desired interactions, each of at most a specified size or show that no such set exists. Under the assumption that there are at most a given number of such interactions causing such a response, locating arrays (LAs) guarantee unique location for every such set of interactions and an algorithm to deal with outliers and nondeterministic behavior from real systems, we additionally require the LAs to have a "separation" between these collections. State-of-the-art approaches generate LAs that can locate at most one interaction of size at most three, due to the massive number of interaction combinations for larger parameters if no constraints are given. This paper presents LocAG, a two-stage algorithm that generates (unconstrained) LAs using a simple, but powerful partitioning strategy of these combinations. In particular, we are able to generate LAs with more factors, with any desired separation, and greater interaction size than existing approaches.
Recent work introduced an algorithm and tool in Coq to automatically repair broken proofs in response to changes that correspond to type equivalences. We report on case studies for manual proof repair across type equivalences using an adaptation of this algorithm in Cubical Agda. Crucially, these case studies capture proof repair use cases that were challenging to impossible in prior work in Coq due to type theoretic limitations, highlighting three benefits to working in Cubical Agda: (1) quotient types enrich the space of repairs we can express as type equivalences, (2) dependent path equality makes it possible to internally state and prove correctness of repaired proofs relative to the original proofs, and (3) functional extensionality and transport make it simple to move between slow and fast computations after repair. They also highlight two challenges of working in Cubical Agda, namely those introduced by: (1) lack of tools for automation, and (2) proof relevance, especially as it interacts with definitional equality. We detail these benefits and challenges in hopes to set the stage for later work in proof repair bridging the benefits of both languages.
While Reinforcement Learning (RL) achieves tremendous success in sequential decision-making problems of many domains, it still faces key challenges of data inefficiency and the lack of interpretability. Interestingly, many researchers have leveraged insights from the causality literature recently, bringing forth flourishing works to unify the merits of causality and address well the challenges from RL. As such, it is of great necessity and significance to collate these Causal Reinforcement Learning (CRL) works, offer a review of CRL methods, and investigate the potential functionality from causality toward RL. In particular, we divide existing CRL approaches into two categories according to whether their causality-based information is given in advance or not. We further analyze each category in terms of the formalization of different models, ranging from the Markov Decision Process (MDP), Partially Observed Markov Decision Process (POMDP), Multi-Arm Bandits (MAB), and Dynamic Treatment Regime (DTR). Moreover, we summarize the evaluation matrices and open sources while we discuss emerging applications, along with promising prospects for the future development of CRL.
Reasoning is a fundamental aspect of human intelligence that plays a crucial role in activities such as problem solving, decision making, and critical thinking. In recent years, large language models (LLMs) have made significant progress in natural language processing, and there is observation that these models may exhibit reasoning abilities when they are sufficiently large. However, it is not yet clear to what extent LLMs are capable of reasoning. This paper provides a comprehensive overview of the current state of knowledge on reasoning in LLMs, including techniques for improving and eliciting reasoning in these models, methods and benchmarks for evaluating reasoning abilities, findings and implications of previous research in this field, and suggestions on future directions. Our aim is to provide a detailed and up-to-date review of this topic and stimulate meaningful discussion and future work.
The demand for artificial intelligence has grown significantly over the last decade and this growth has been fueled by advances in machine learning techniques and the ability to leverage hardware acceleration. However, in order to increase the quality of predictions and render machine learning solutions feasible for more complex applications, a substantial amount of training data is required. Although small machine learning models can be trained with modest amounts of data, the input for training larger models such as neural networks grows exponentially with the number of parameters. Since the demand for processing training data has outpaced the increase in computation power of computing machinery, there is a need for distributing the machine learning workload across multiple machines, and turning the centralized into a distributed system. These distributed systems present new challenges, first and foremost the efficient parallelization of the training process and the creation of a coherent model. This article provides an extensive overview of the current state-of-the-art in the field by outlining the challenges and opportunities of distributed machine learning over conventional (centralized) machine learning, discussing the techniques used for distributed machine learning, and providing an overview of the systems that are available.
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191