In light of the rapidly evolving capabilities of large language models (LLMs), it becomes imperative to develop rigorous domain-specific evaluation benchmarks to accurately assess their capabilities. In response to this need, this paper introduces ArcMMLU, a specialized benchmark tailored for the Library & Information Science (LIS) domain in Chinese. This benchmark aims to measure the knowledge and reasoning capability of LLMs within four key sub-domains: Archival Science, Data Science, Library Science, and Information Science. Following the format of MMLU/CMMLU, we collected over 6,000 high-quality questions for the compilation of ArcMMLU. This extensive compilation can reflect the diverse nature of the LIS domain and offer a robust foundation for LLM evaluation. Our comprehensive evaluation reveals that while most mainstream LLMs achieve an average accuracy rate above 50% on ArcMMLU, there remains a notable performance gap, suggesting substantial headroom for refinement in LLM capabilities within the LIS domain. Further analysis explores the effectiveness of few-shot examples on model performance and highlights challenging questions where models consistently underperform, providing valuable insights for targeted improvements. ArcMMLU fills a critical gap in LLM evaluations within the Chinese LIS domain and paves the way for future development of LLMs tailored to this specialized area.

Human behavior is inherently hierarchical, resulting from the decomposition of a task into subtasks or an abstract action into concrete actions. However, behavior is typically measured as a sequence of actions, which makes it difficult to infer its hierarchical structure. In this paper, we explore how people form hierarchically-structured plans, using an experimental paradigm that makes hierarchical representations observable: participants create programs that produce sequences of actions in a language with explicit hierarchical structure. This task lets us test two well-established principles of human behavior: utility maximization (i.e. using fewer actions) and minimum description length (MDL; i.e. having a shorter program). We find that humans are sensitive to both metrics, but that both accounts fail to predict a qualitative feature of human-created programs, namely that people prefer programs with reuse over and above the predictions of MDL. We formalize this preference for reuse by extending the MDL account into a generative model over programs, modeling hierarchy choice as the induction of a grammar over actions. Our account can explain the preference for reuse and provides the best prediction of human behavior, going beyond simple accounts of compressibility to highlight a principle that guides hierarchical planning.

In recent years, learning-based approaches have demonstrated significant promise in addressing intricate navigation tasks. Traditional methods for training deep neural network navigation policies rely on meticulously designed reward functions or extensive teleoperation datasets as navigation demonstrations. However, the former is often confined to simulated environments, and the latter demands substantial human labor, making it a time-consuming process. Our vision is for robots to autonomously learn navigation skills and adapt their behaviors to environmental changes without any human intervention. In this work, we discuss the self-supervised navigation problem and present Dynamic Graph Memory (DGMem), which facilitates training only with on-board observations. With the help of DGMem, agents can actively explore their surroundings, autonomously acquiring a comprehensive navigation policy in a data-efficient manner without external feedback. Our method is evaluated in photorealistic 3D indoor scenes, and empirical studies demonstrate the effectiveness of DGMem.

People in conversation entrain their linguistic behaviours through spontaneous alignment mechanisms [7] - both in face-to-face and computer-mediated communication (CMC) [8]. In CMC, one of the mechanisms through which linguistic entrainment happens is through explicit replies. Indeed, the use of explicit replies influences the structure of conversations, favouring the formation of reply-trees typically delineated by topic shifts [5]. The interpersonal coordination mechanisms realized by how actors address each other have been studied using a probabilistic framework proposed by David Gibson [2,3]. Other recent approaches use computational methods and information theory to quantify changes in text. We explore coordination mechanisms concerned with some of the roles utterances play in dialogues - specifically in explicit replies. We identify these roles by finding community structure in the conversation's vocabulary using a non-parametric, hierarchical topic model. Some conversations may always stay on the ground, remaining at the level of general introductory chatter. Some others may develop a specific sub-topic in significant depth and detail. Even others may jump between general chatter, out-of-topic remarks and people agreeing or disagreeing without further elaboration.

To precisely evaluate a language model's capability for logical reading comprehension, we present a dataset for testing the understanding of the rationale behind critical reasoning. For questions taken from an existing multiplechoice logical reading comprehension dataset, we crowdsource rationale texts that explain why we should select or eliminate answer options, resulting in 3,003 multiple-choice subquestions that are associated with 943 main questions. Experiments on our dataset show that recent large language models (e.g., InstructGPT) struggle to answer the subquestions even if they are able to answer the main questions correctly. We find that the models perform particularly poorly in answering subquestions written for the incorrect options of the main questions, implying that the models have a limited capability for explaining why incorrect alternatives should be eliminated. These results suggest that our dataset encourages further investigation into the critical reasoning ability of language models while focusing on the elimination process of relevant alternatives.

The advent of artificial intelligence (AI) has instigated a transformational wave across various sectors, with education standing as a salient beneficiary. AI's unrivaled capacity to enable personalized and adaptive learning experiences has propelled intelligent tutoring systems to the forefront of modern educational paradigms (Kasneci et al., 2023). These systems, powered by AI, offer individualized feedback and interactive learning modules designed to cater to each student's distinct learning needs. Nonetheless, the challenge of developing AI tutors capable of delivering consistently accurate and dependable responses across diverse academic disciplines persists. A notable hindrance to the reliability of AI in educational applications is the occurrence of'information hallucination', a phenomenon where AI-generated responses, while appearing valid, deviate from factual accuracy (Nye et al., 2023). Such inconsistencies can undermine confidence in AI-centric educational systems (Kasneci et al., 2023). Furthermore, the customization of these systems to align with specific course content necessitates access to current and pertinent educational materials, a task often complicated by the multifaceted nature of academic disciplines. To tackle these challenges, this paper introduces AI Tutor, a web application developed upon the sophisticated infrastructure of large language models (LLMs) and retrieval-augmented generation (RAG). AI Tutor is engineered to deliver accurate, contextually relevant responses by intelligently assimilating information from course-specific materials (Lewis et al., 2020).

Drawing inspiration from prompt tuning techniques applied to Large Language Models, recent methods based on pre-trained ViT networks have achieved remarkable results in the field of Continual Learning. Specifically, these approaches propose to maintain a set of prompts and allocate a subset of them to learn each task using a key-query matching strategy. However, they may encounter limitations when lacking control over the correlations between old task queries and keys of future tasks, the shift of features in the latent space, and the relative separation of latent vectors learned in independent tasks. In this work, we introduce a novel key-query learning strategy based on orthogonal projection, inspired by model-agnostic meta-learning, to enhance prompt matching efficiency and address the challenge of shifting features. Furthermore, we introduce a One-Versus-All (OVA) prototype-based component that enhances the classification head distinction. Experimental results on benchmark datasets demonstrate that our method empowers the model to achieve results surpassing those of current state-of-the-art approaches by a large margin of up to 20%. Our code is available at https://anonymous.4open.science/r/KOPPA/README.md. Continual Learning (CL) is an evolving field in machine learning, aiming to enable models to learn continuously from a sequence of tasks with varying data distributions. A challenging CL scenario is Class Incremental Learning (CIL), where a model sequentially learns new categories and must classify all seen classes without task-ID information, leading to a fundamental issue in CL known as Catastrophic Forgetting (CF) (French, 1999), where performance on earlier tasks degrades due to the absence of old task data and differences in data distributions. In CIL, models are required to classify test samples without prior knowledge of their task IDs.

We present Llemma, a large language model for mathematics. We continue pretraining Code Llama on the Proof-Pile-2, a mixture of scientific papers, web data containing mathematics, and mathematical code, yielding Llemma. On the MATH benchmark Llemma outperforms all known open base models, as well as the unreleased Minerva model suite on an equi-parameter basis. Moreover, Llemma is capable of tool use and formal theorem proving without any further finetuning. We openly release all artifacts, including 7 billion and 34 billion parameter models, the Proof-Pile-2, and code to replicate our experiments.

Though LLMs are capable of generating plausible programs, it's challenging to interact with the LLMs further to revise the program, especially if the user's specific requirements are different from the initial proposal. In this paper, we introduce ANPL, an interactive programming system that ensures users can always refine the generated code towards their specific programmatic intents via structured decompositions. Borrowing the paradigm of sketching from program synthesis, an ANPL program consists of a set of input-outputs that it must satisfy, a ``sketch'' -- control/data flow expressed in precise code (e.g. Python), and ``holes'' -- sub-modules to be implemented by the LLM specified with natural language. The user revises an ANPL program by either modifying the sketch, changing the language used to describe the holes, or providing additional input-outputs to a particular hole, turning it into a sub-ANPL program that can be solved recursively. This workflow allows the users to offload programming burdens to the LLM as much as possible while retaining the ability to pinpoint and resolve bugs locally, without exposing the rest of the program to the LLM. We deploy ANPL on the Abstraction and Reasoning Corpus (ARC), a set of unique tasks that are challenging for state-of-the-art AI systems, showing it outperforms baseline programming systems that (a) without the ability to decompose tasks interactively and (b) without the guarantee that the modules can be correctly composed together. Additional evaluations on APPS, HumanEval, and real-world programming tasks have validated that the ANPL framework is applicable to multiple programming domains. We release the ANPL solutions to the ARC tasks as a dataset, providing insights into how humans decompose novel tasks programmatically. See our code at https://iprc-dip.github.io/ANPL/.

This paper presents an experiment of automatically scoring handwritten descriptive answers in the trial tests for the new Japanese university entrance examination, which were made for about 120,000 examinees in 2017 and 2018. There are about 400,000 answers with more than 20 million characters. Although all answers have been scored by human examiners, handwritten characters are not labeled. We present our attempt to adapt deep neural network-based handwriting recognizers trained on a labeled handwriting dataset into this unlabeled answer set. Our proposed method combines different training strategies, ensembles multiple recognizers, and uses a language model built from a large general corpus to avoid overfitting into specific data. In our experiment, the proposed method records character accuracy of over 97% using about 2,000 verified labeled answers that account for less than 0.5% of the dataset. Then, the recognized answers are fed into a pre-trained automatic scoring system based on the BERT model without correcting misrecognized characters and providing rubric annotations. The automatic scoring system achieves from 0.84 to 0.98 of Quadratic Weighted Kappa (QWK). As QWK is over 0.8, it represents an acceptable similarity of scoring between the automatic scoring system and the human examiners. These results are promising for further research on end-to-end automatic scoring of descriptive answers.