Chain-of-Thought (CoT) prompting has boosted the multi-step reasoning capabilities of Large Language Models (LLMs) by generating a series of rationales before the final answer. We analyze the reasoning paths generated by CoT and find two issues in multi-step reasoning: (i) Generating rationales irrelevant to the question, (ii) Unable to compose subquestions or queries for generating/retrieving all the relevant information. To address them, we propose a graph-guided CoT prompting method, which guides the LLMs to reach the correct answer with graph representation/verification steps. Specifically, we first leverage LLMs to construct a "question/rationale graph" by using knowledge extraction prompting given the initial question and the rationales generated in the previous steps. Then, the graph verification step diagnoses the current rationale triplet by comparing it with the existing question/rationale graph to filter out irrelevant rationales and generate follow-up questions to obtain relevant information. Additionally, we generate CoT paths that exclude the extracted graph information to represent the context information missed from the graph extraction. Our graph-guided reasoning method shows superior performance compared to previous CoT prompting and the variants on multi-hop question answering benchmark datasets.

The versatility of Large Language Models (LLMs) on natural language understanding tasks has made them popular for research in social sciences. In particular, to properly understand the properties and innate personas of LLMs, researchers have performed studies that involve using prompts in the form of questions that ask LLMs of particular opinions. In this study, we take a cautionary step back and examine whether the current format of prompting enables LLMs to provide responses in a consistent and robust manner. We first construct a dataset that contains 693 questions encompassing 39 different instruments of persona measurement on 115 persona axes. Additionally, we design a set of prompts containing minor variations and examine LLM's capabilities to generate accurate answers, as well as consistency variations to examine their consistency towards simple perturbations such as switching the option order. Our experiments on 15 different open-source LLMs reveal that even simple perturbations are sufficient to significantly downgrade a model's question-answering ability, and that most LLMs have low negation consistency. Our results suggest that the currently widespread practice of prompting is insufficient to accurately capture model perceptions, and we discuss potential alternatives to improve such issues.

Large language model (LLM) inference is computation and memory intensive, so we adapt lexical shortlisting to it hoping to improve both. While lexical shortlisting is well-explored in tasks like machine translation, it requires modifications before being suitable for LLMs as the intended applications vary significantly. Our work studies two heuristics to shortlist sub-vocabulary at LLM inference time: Unicode-based script filtering and corpus-based selection. We explore different LLM families and sizes, and we find that lexical shortlisting can reduce the memory usage of some models by nearly 50\% and has an upper bound of 25\% improvement in generation speed. In this pilot study, we also identify the drawbacks of such vocabulary selection methods and propose avenues for future research.

Research automation efforts usually employ AI as a tool to automate specific tasks within the research process. To create an AI that truly conduct research themselves, it must independently generate hypotheses, design verification plans, and execute verification. Therefore, we investigated if an AI itself could autonomously generate and verify hypothesis for a toy machine learning research problem. We prompted GPT-4 to generate hypotheses and Python code for hypothesis verification with limited methodological guidance. Our findings suggest that, in some instances, GPT-4 can autonomously generate and validate hypotheses without detailed guidance. While this is a promising result, we also found that none of the verifications were flawless, and there remain significant challenges in achieving autonomous, human-level research using only generic instructions. These findings underscore the need for continued exploration to develop a general and autonomous AI researcher.

Recently, ChatGPT has emerged as a powerful NLP tool that can carry out several tasks. However, the range of languages ChatGPT can handle remains largely a mystery. In this work, we investigate ChatGPT's language identification abilities. For this purpose, we compile Babel-670, a benchmark comprising $670$ languages representing $23$ language families. Languages in Babel-670 run the gamut between the very high-resource to the very low-resource and are spoken in five continents. We then study ChatGPT's (both GPT-3.5 and GPT-4) ability to (i) identify both language names and language codes (ii) under both zero- and few-shot conditions (iii) with and without provision of label set. When compared to smaller finetuned language identification tools, we find that ChatGPT lags behind. Our empirical analysis shows the reality that ChatGPT still resides in a state of potential enhancement before it can sufficiently serve diverse communities.

People share stories online for a myriad of purposes, whether as a means of self-disclosure, processing difficult personal experiences, providing needed information or entertainment, or persuading others to share their beliefs. Better understanding of online storytelling can illuminate the dynamics of social movements, sensemaking practices, persuasion strategies, and more. However, unlike other media such as books and visual content where the narrative nature of the content is often overtly signaled at the document level, studying storytelling in online communities is challenging due to the mixture of storytelling and non-storytelling behavior, which can be interspersed within documents and across diverse topics and settings. We introduce a codebook and create the Storytelling in Online Communities Corpus, an expert-annotated dataset of 502 English-language posts and comments with labeled story and event spans. Using our corpus, we train and evaluate an online story detection model, which we use to investigate the role storytelling of in different social contexts. We identify distinctive features of online storytelling, the prevalence of storytelling among different communities, and the conversational patterns of storytelling.

One of the fundamental skills required for an agent acting in an environment to complete tasks is the ability to understand what actions are plausible at any given point. This work explores a novel use of code representations to reason about action preconditions for sequential decision making tasks. Code representations offer the flexibility to model procedural activities and associated constraints as well as the ability to execute and verify constraint satisfaction. Leveraging code representations, we extract action preconditions from demonstration trajectories in a zero-shot manner using pre-trained code models. Given these extracted preconditions, we propose a precondition-aware action sampling strategy that ensures actions predicted by a policy are consistent with preconditions. We demonstrate that the proposed approach enhances the performance of few-shot policy learning approaches across task-oriented dialog and embodied textworld benchmarks.

Recent studies have found that summaries generated by large language models (LLMs) are favored by human annotators over the original reference summaries in commonly used summarization datasets. Therefore, we investigate a new learning setting of text summarization models that considers the LLMs as the reference or the gold-standard oracle on these datasets. To examine the standard practices that are aligned with this new learning setting, we investigate two LLM-based summary quality evaluation methods for model training and adopt a contrastive learning training method to leverage the LLM-guided learning signals. Our experiments on the CNN/DailyMail and XSum datasets demonstrate that smaller summarization models can achieve similar performance as LLMs under LLM-based evaluation. However, we found that the smaller models can not yet reach LLM-level performance under human evaluation despite promising improvements brought by our proposed training methods. Meanwhile, we perform a meta-analysis on this new learning setting that reveals a discrepancy between human and LLM-based evaluation, highlighting the benefits and risks of this LLM-as-reference setting we investigated.

A popular approach for improving the correctness of output from large language models (LLMs) is Self-Consistency - poll the LLM multiple times and output the most frequent solution. Existing Self-Consistency techniques always generate a constant number of samples per question, where a better approach will be to non-uniformly distribute the available budget based on the amount of agreement in the samples generated so far. In response, we introduce Adaptive-Consistency, a cost-efficient, model-agnostic technique that dynamically adjusts the number of samples per question using a lightweight stopping criterion. Our experiments over 17 reasoning and code generation datasets and three LLMs demonstrate that Adaptive-Consistency reduces sample budget by up to 7.9 times with an average accuracy drop of less than 0.1%. Our code and data are available at https://www.sample-step-by-step.info

Large language models (LLMs) like GPT are often conceptualized as passive predictors, simulators, or even stochastic parrots. We instead conceptualize LLMs by drawing on the theory of active inference originating in cognitive science and neuroscience. We examine similarities and differences between traditional active inference systems and LLMs, leading to the conclusion that, currently, LLMs lack a tight feedback loop between acting in the world and perceiving the impacts of their actions, but otherwise fit in the active inference paradigm. We list reasons why this loop may soon be closed, and possible consequences of this including enhanced model self-awareness and the drive to minimize prediction error by changing the world.