Language models are increasingly being deployed for general problem solving across a wide range of tasks, but are still confined to token-level, left-to-right decision-making processes during inference. This means they can fall short in tasks that require exploration, strategic lookahead, or where initial decisions play a pivotal role. To surmount these challenges, we introduce a new framework for language model inference, Tree of Thoughts (ToT), which generalizes over the popular Chain of Thought approach to prompting language models, and enables exploration over coherent units of text (thoughts) that serve as intermediate steps toward problem solving. ToT allows LMs to perform deliberate decision making by considering multiple different reasoning paths and self-evaluating choices to decide the next course of action, as well as looking ahead or backtracking when necessary to make global choices.Our experiments show that ToT significantly enhances language models' problem-solving abilities on three novel tasks requiring non-trivial planning or search: Game of 24, Creative Writing, and Mini Crosswords. For instance, in Game of 24, while GPT-4 with chain-of-thought prompting only solved 4\% of tasks, our method achieved a success rate of 74\%.

Given data, finding a faithful low-dimensional hyperbolic embedding of the data is a key method by which we can extract hierarchical information or learn representative geometric features of the data. In this paper, we explore a new method for learning hyperbolic representations by taking a metric-first approach. Rather than determining the low-dimensional hyperbolic embedding directly, we learn a tree structure on the data. This tree structure can then be used directly to extract hierarchical information, embedded into a hyperbolic manifold using Sarkar's construction \cite{sarkar}, or used as a tree approximation of the original metric. To this end, we present a novel fast algorithm \textsc{TreeRep} such that, given a $\delta$-hyperbolic metric (for any $\delta \geq 0$), the algorithm learns a tree structure that approximates the original metric. In the case when $\delta = 0$, we show analytically that \textsc{TreeRep} exactly recovers the original tree structure. We show empirically that \textsc{TreeRep} is not only many orders of magnitude faster than previously known algorithms, but also produces metrics with lower average distortion and higher mean average precision than most previous algorithms for learning hyperbolic embeddings, extracting hierarchical information, and approximating metrics via tree metrics.

Languages differ in how they divide up the world into concepts and words; e.g., in contrast to English, Swahili has a single concept for `belly' and `womb'. We investigate these differences in conceptualization across 1,335 languages by aligning concepts in a parallel corpus. To this end, we propose Conceptualizer, a method that creates a bipartite directed alignment graph between source language concepts and sets of target language strings. In a detailed linguistic analysis across all languages for one concept (`bird') and an evaluation on gold standard data for 32 Swadesh concepts, we show that Conceptualizer has good alignment accuracy. We demonstrate the potential of research on conceptualization in NLP with two experiments. (1) We define crosslingual stability of a concept as the degree to which it has 1-1 correspondences across languages, and show that concreteness predicts stability. (2) We represent each language by its conceptualization pattern for 83 concepts, and define a similarity measure on these representations. The resulting measure for the conceptual similarity of two languages is complementary to standard genealogical, typological, and surface similarity measures. For four out of six language families, we can assign languages to their correct family based on conceptual similarity with accuracy between 54% and 87%.

One of the biggest confusions about "Artificial Intelligence" is that it is a very vague term. That's because Artificial Intelligence or AI is a term that was coined way back in 1955 with extreme hubris: We propose that a 2 month, 10 man study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire. The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it. An attempt will be made to find how to make machines use language, form abstractions, and concepts, solve kinds of problems now reserved for humans, and improve themselves. AI is over half a century old and carries with it too much baggage.

In this guide, we'll take a practical, concise tour through modern machine learning algorithms. While other such lists exist, they don't really explain the practical tradeoffs of each algorithm, which we hope to do here. We'll discuss the advantages and disadvantages of each algorithm based on our experience. Categorizing machine learning algorithms is tricky, and there are several reasonable approaches; they can be grouped into generative/discriminative, parametric/non-parametric, supervised/unsupervised, and so on. However, from our experience, this isn't always the most practical way to group algorithms.

Modeling various aspects of language--syntax, semantics, pragmatics, and discourse, among others--by the use of constrained formal-computational systems, just adequate for such modeling, has proved to be an effective research strategy, leading to deep understanding of these aspects, with implications for both machine processing and human processing. This approach enables one to distinguish between the universal and stipulative constraints.

Department of Computer Science Rutgers Universaty New Brunswick, New Jersey 08903 Abstract In this article we discuss a method for learning useful conditions on the application of operators during heuristic search Since learning is not attempted until a complete solution path has been found for a problem, credit for correct moves and blame for incorrect moves is easily assigned We review four learning systems that have incorporated similar techniques to learn in the domains of algebra, symbolic integration, and puzzle-solving We conclude that the basic approach of learning from solution paths can be applied t,o any situation in which problems can be solved by sequential search Finally, we examine some potential difficulties that may arise in more complex domains, and suggest some possible extensions for dealing with them. PEOPLE LEARN FROM EXPERIENCE, and for the past 25 years, Artificial Intelligence researchers have been attempting to replicate this process. In t,his article we focus on learning in domains where search is involved. Furthermore, we will restrict our attention t,o cases in which the legal operators for a task are known, and the learning task is to determine the conditions under which those operators can be usefully applied. Once such a set of heuristically useful conditions has been discovered, search will be directed down profitable We would like to thank Jaime Carbonell and Hans Berliner for helpful comments on an earlier version of this article.

Empirical evidence suggests that searching deeper in game trees using the minimax propagation rule usually improves the quality of decisions significantly. However, despite many recent theoretical analyses of the effects of minimax look-ahead, however, this phenomenon has still not been convincingly explained. Instead, much attention has been given to socalled pathological behavior, which occurs under certain assumptions. This article supports the view that pathology is a direct result of these underlying theoretical assumptions. Pathology does not occur in practice, because these assumptions do not apply in realistic domains.

I will review the state of solving incomplete-information games. They encompass many practical problems such as auctions, negotiations, and security applications. I will discuss them in the context of how they have transformed computer poker. In short, game-theoretic reasoning now scales to many large problems, outperforms the alternatives on those problems, and in some games beats the best humans. This is nontrivial because an agent's utility-maximizing strategy generally depends on the other agents' strategies.

A more modern view is to envision drivers and passengers as actively interacting with a complex automated system. Such interactive activity leads us to consider intelligent and advanced ways of interaction leading to cars that can adapt to their drivers. In this article, we focus on the adaptive cruise control (ACC) technology that allows a vehicle to automatically adjust its speed to maintain a preset distance from the vehicle in front of it based on the driver's preferences. Although individual drivers have different driving styles and preferences, current systems do not distinguish among users. We introduce a method to combine machine-learning algorithms with demographic information and expert advice into existing automated assistive systems.