ACL 2024 Papers

People tend to use language to mention surprising properties of events: for example, when a banana is blue, we are more likely to mention color than when it is yellow. This fact is taken to suggest that yellowness is somehow a typical feature of bananas, and blueness is exceptional. Similar to how a yellow color is typical of bananas, there may also be genders that are typical of occupations. In this work, we explore this question using information theoretic techniques coupled with corpus statistic analysis. In two distinct large corpora, we do not find strong evidence that occupations and gender display the same patterns of mentioning as do bananas and color. Instead, we find that gender mentioning is correlated with femaleness of occupation in particular, suggesting perhaps that woman-dominated occupations are seen as somehow “more gendered” than male-dominated ones, and thereby they encourage more gender mentioning overall.

Large language models, comprising billions of parameters and pre-trained on extensive web-scale corpora, have been claimed to acquire certain capabilities without having been specifically trained on them. These capabilities, referred to as “emergent abilities,” have been a driving force in discussions regarding the potentials and risks of language models. A key challenge in evaluating emergent abilities is that they are confounded by model competencies that arise through alternative prompting techniques, including in-context learning, which is the ability of models to complete a task based on a few examples. We present a novel theory that explains emergent abilities, taking into account their potential confounding factors, and rigorously substantiate this theory through over 1000 experiments. Our findings suggest that purported emergent abilities are not truly emergent, but result from a combination of in-context learning, model memory, and linguistic knowledge. Our work is a foundational step in explaining language model performance, providing a template for their efficient use and clarifying the paradox of their ability to excel in some instances while faltering in others. Thus, we demonstrate that their capabilities should not be overestimated.

The natural language processing field has been evolving around language models for the past few years, from the usage of n-gram language models for re-ranking, to transfer learning with encoder-only (BERT-like) language models, and finally to large language models (LLMs) as general solvers. LLMs are dominated by the decoder-only type, and they are popular for their efficacy in numerous tasks. LLMs are regarded as having strong comprehension abilities and strong capabilities to solve new unseen tasks. As such, people may quickly assume that decoder-only LLMs always perform better than the encoder-only ones, especially for understanding word meaning. In this paper, we demonstrate that decoder-only LLMs perform worse on word meaning comprehension than an encoder-only language model that has vastly fewer parameters.

The widespread use of large language models (LLMs) has sparked concerns about the potential misuse of AI-generated text, as these models can produce content that closely resembles human-generated text. Current detectors for AI-generated text (AIGT) lack robustness against adversarial perturbations, with even minor changes in characters or words causing a reversal in distinguishing between human-created and AI-generated text. This paper investigates the robustness of existing AIGT detection methods and introduces a novel detector, the Siamese Calibrated Reconstruction Network (SCRN). The SCRN employs a reconstruction network to add and remove noise from text, extracting a semantic representation that is robust to local perturbations. We also propose a siamese calibration technique to train the model to make equally confident predictions under different noise, which improves the model’s robustness against adversarial perturbations. Experiments on four publicly available datasets show that the SCRN outperforms all baseline methods, achieving 6.5%-18.25% absolute accuracy improvement over the best baseline method under adversarial attacks. Moreover, it exhibits superior generalizability in cross-domain, cross-genre, and mixed-source scenarios. The code is available at https://github.com/CarlanLark/Robust-AIGC-Detector.

This paper aims to extend the code generation capability of large language models (LLMs) to automatically manage comprehensive software requirements from given textual descriptions. Such requirements include both functional (i.e. achieving expected behavior for inputs) and non-functional (e.g., time/space performance, robustness, maintainability) requirements. However, textual descriptions can either express requirements verbosely or may even omit some of them. We introduce ARCHCODE, a novel framework that leverages in-context learning to organize requirements observed in descriptions and to extrapolate unexpressed requirements from them. ARCHCODE generates requirements from given descriptions, conditioning them to produce code snippets and test cases. Each test case is tailored to one of the requirements, allowing for the ranking of code snippets based on the compliance of their execution results with the requirements. Public benchmarks show that ARCHCODE enhances to satisfy functional requirements, significantly improving Pass@k scores.Furthermore, we introduce HumanEval-NFR, the first evaluation of LLMs’ non-functional requirements in code generation, demonstrating ARCHCODE’s superiority over baseline methods. The implementation of ARCHCODE and the HumanEval-NFR benchmark are both publicly accessible.

The focus of language model evaluation has transitioned towards reasoning and knowledge-intensive tasks, driven by advancements in pretraining large models. While state-of-the-art models are partially trained on large Arabic texts, evaluating their performance in Arabic remains challenging due to the limited availability of relevant datasets. To bridge this gap, we present ArabicMMLU, the first multi-task language understanding benchmark for the Arabic language, sourced from school exams across diverse educational levels in different countries spanning North Africa, the Levant, and the Gulf regions. Our data comprises 40 tasks and 14,575 multiple-choice questions in Modern Standard Arabic (MSA) and is carefully constructed by collaborating with native speakers in the region. Our comprehensive evaluations of 35 models reveal substantial room for improvement, particularly among the best open-source models. Notably, BLOOMZ, mT0, LLama2, and Falcon struggle to achieve a score of 50%, while even the top-performing Arabic-centric model only achieves a score of 62.3%.

The widespread absence of diacritical marks in Arabic text poses a significant challenge for Arabic natural language processing (NLP). This paper explores instances of naturally occurring diacritics, referred to as “diacritics in the wild,” to unveil patterns and latent information across six diverse genres: news articles, novels, children’s books, poetry, political documents, and ChatGPT outputs. We present a new annotated dataset that maps real-world partially diacritized words to their maximal full diacritization in context. Additionally, we propose extensions to the analyze-and-disambiguate approach in Arabic NLP to leverage these diacritics, resulting in notable improvements. Our contributions encompass a thorough analysis, valuable datasets, and an extended diacritization algorithm. We release our code and datasets as open source.

Autonomous agents that address day-to-day digital tasks (e.g., ordering groceries for a household), must not only operate multiple apps (e.g., notes, messaging, shopping app) via APIs, but also generate rich code with complex control flow in an iterative manner based on their interaction with the environment. However, existing benchmarks for tool use are inadequate, as they only cover tasks that require a simple sequence of API calls. To remedy this gap, we built AppWorld Engine, a high-quality execution environment (60K lines of code) of 9 day-to-day apps operable via 457 APIs and populated with realistic digital activities simulating the lives of ~100 fictitious users. We then created AppWorld Benchmark (40K lines of code), a suite of 750 natural, diverse, and challenging autonomous agent tasks requiring rich and interactive code generation. It supports robust programmatic evaluation with state-based unit tests, allowing for different ways of completing a task while also checking for unexpected changes, i.e., collateral damage. The state-of-the-art LLM, GPT4O, solves only ~49% of our ‘normal’ tasks and ~30% of ‘challenge’ tasks, while other models solve at least 16% fewer. This highlights the benchmark’s difficulty and AppWorld’s potential to push the frontiers of interactive coding agents.

Text embedding is pivotal in semantic textual similarity (STS) tasks, which are crucial components in Large Language Model (LLM) applications. STS learning largely relies on the cosine function as the optimization objective to reflect semantic similarity. However, the cosine has saturation zones rendering vanishing gradients and hindering learning subtle semantic differences in text embeddings. To address this issue, we propose a novel Angle-optimized Embedding model, AoE. It optimizes angle differences in complex space to explore similarity in saturation zones better. To set up a comprehensive evaluation, we experimented with existing short-text STS, our newly collected long-text STS, and downstream task datasets. Extensive experimental results on STS and MTEB benchmarks show that AoE significantly outperforms popular text embedding models neglecting cosine saturation zones. It highlights that AoE can produce high-quality text embeddings and broadly benefit downstream tasks.

We introduce AnyGPT, an any-to-any multimodal language model that utilizes discrete representations for the unified processing of various modalities, including speech, text, images, and music. AnyGPT can be trained stably without any alterations to the current large language model (LLM) architecture or training paradigms. Instead, it relies exclusively on data-level preprocessing, facilitating the seamless integration of new modalities into LLMs, akin to the incorporation of new languages.We build a multimodal text-centric dataset for multimodal alignment pre-training. Utilizing generative models, we synthesize the first large-scale any-to-any multimodal instruction dataset. It consists of 108k samples of multi-turn conversations that intricately interweave various modalities, thus equipping the model to handle arbitrary combinations of multimodal inputs and outputs.Experimental results demonstrate that AnyGPT is capable of facilitating any-to-any multimodal conversation while achieving performance comparable to specialized models across all modalities, proving that discrete representations can effectively and conveniently unify multiple modalities within a language model. Demos are shown in https://junzhan2000.github.io/AnyGPT.github.io/.

This work aims to build a text embedder that can capture characteristics of texts specified by user instructions clarifying the similarity criterion. While previous methods improve general task awareness by injecting the instruction information into encoding, they fail to be sensitive to clearer criteria like “evaluate similarity based on emotion”. We instead propose a different viewpoint, which treats the instruction as a “question” about the input text and encodes the expected answers to obtain the representation accordingly. Intuitively, texts with the same (implicit) semantics would share similar answers following the instruction, thus leading to more similar representations. Specifically, we propose InBedder that instantiates this learning-to-answer idea by only fine-tuning language models via abstractive question answering tasks. Despite its simplicity, InBedder demonstrates significantly improved instruction-following capabilities according to our proposed instruction awareness tests and instruction robustness tests, when applied to language models with large language models (LLMs) (e.g., llama-2-7b) and smaller encoder-based LMs (e.g., roberta-large). Additionally, our qualitative analysis of clustering outcomes, achieved by applying diverse instructions to the same unlabeled corpus, demonstrates a high degree of interpretability in the clusters formed.

Despite the remarkable generative capabilities of language models in producing naturalistic language, their effectiveness on explicit manipulation and generation of linguistic structures remain understudied. In this paper, we investigate the task of generating new sentences preserving a given semantic structure, following the FrameNet formalism. We propose a framework to produce novel frame-semantically annotated sentences following an overgenerate-and-filter approach. Our results show that conditioning on rich, explicit semantic information tends to produce generations with high human acceptance, under both prompting and finetuning. Our generated frame-semantic structured annotations are effective at training data augmentation for frame-semantic role labeling in low-resource settings; however, we do not see benefits under higher resource settings. Our study concludes that while generating high-quality, semantically rich data might be within reach, the downstream utility of such generations remains to be seen, highlighting the outstanding challenges with automating linguistic annotation tasks.

Large language models (LLMs) reflect societal norms and biases, especially about gender. While societal biases and stereotypes have been extensively researched in various NLP applications, there is a surprising gap for emotion analysis. However, emotion and gender are closely linked in societal discourse. E.g., women are often thought of as more empathetic, while men’s anger is more socially accepted. To fill this gap, we present the first comprehensive study of gendered emotion attribution in five state-of-the-art LLMs (open- and closed-source). We investigate whether emotions are gendered, and whether these variations are based on societal stereotypes. We prompt the models to adopt a gendered persona and attribute emotions to an event like ‘When I had a serious argument with a dear person’. We then analyze the emotions generated by the models in relation to the gender-event pairs. We find that all models consistently exhibit gendered emotions, influenced by gender stereotypes. These findings are in line with established research in psychology and gender studies. Our study sheds light on the complex societal interplay between language, gender, and emotion. The reproduction of emotion stereotypes in LLMs allows us to use those models to study the topic in detail, but raises questions about the predictive use of those same LLMs for emotion applications.

The ancestor of Chinese character – the ancient characters from about 1300 BC to 200 BC are not fixed in their writing glyphs. At the same or different points in time, one character can possess multiple glyphs that are different in shapes or radicals. Nearly half of ancient glyphs have not been deciphered yet. This paper proposes an innovative task of ancient Chinese glyph identification, which aims at inferring the Chinese character label for the unknown ancient Chinese glyphs which are not in the training set based on the image and radical information. Specifically, we construct a Chinese glyph knowledge graph (CGKG) associating glyphs in different historical periods according to the radical semantics, and propose a multimodal Chinese glyph identification framework (MCGI) fusing the visual, textual, and the graph data. The experiment is designed on a real Chinese glyph dataset spanning over 1000 years, it demonstrates the effectiveness of our method, and reports the potentials of each modality on this task. It provides a preliminary reference for the automatic ancient Chinese character deciphering at the glyph level.

Large language models (LLMs) predominantly employ decoder-only transformer architectures, necessitating the retention of keys/values information for historical tokens to provide contextual information and avoid redundant computation. However, the substantial size and parameter volume of these LLMs require massive GPU memory. This memory demand increases with the length of the input text, leading to an urgent need for more efficient methods of information storage and processing. This study introduces Anchor-based LLMs (AnLLMs), which utilize an innovative anchor-based self-attention network (AnSAN) and also an anchor-based inference strategy. This approach enables LLMs to compress sequence information into an anchor token, reducing the keys/values cache and enhancing inference efficiency. Experiments on question-answering benchmarks reveal that AnLLMs maintain similar accuracy levels while achieving up to 99% keys/values cache reduction and up to 3.5 times faster inference. Despite a minor compromise in accuracy, the substantial enhancements of AnLLMs employing the AnSAN technique in resource utilization and computational efficiency underscore their potential for practical LLM applications.

The digital landscape is rapidly evolving with an ever-increasing volume of online news, emphasizing the need for swift and precise analysis of complex events.We refer to the complex events composed of many news articles over an extended period as Temporal Complex Event (TCE). This paper proposes a novel approach using Large Language Models (LLMs) to systematically extract and analyze the event chain within TCE, characterized by their key points and timestamps. We establish a benchmark, named TCELongBench, to evaluate the proficiency of LLMs in handling temporal dynamics and understanding extensive text. This benchmark encompasses three distinct tasks - reading comprehension, temporal sequencing, and future event forecasting. In the experiment, we leverage retrieval-augmented generation (RAG) method and LLMs with long context window to deal with lengthy news articles of TCE. Our findings indicate that models with suitable retrievers exhibit comparable performance with those utilizing long context window.

Modern language models are capable of contextualizing words based on their surrounding context. However, this capability is often compromised due to semantic change that leads to words being used in new, unexpected contexts not encountered during pre-training. In this paper, we model semantic change by studying the effect of unexpected contexts introduced by lexical replacements. We propose a replacement schema where a target word is substituted with lexical replacements of varying relatedness, thus simulating different kinds of semantic change. Furthermore, we leverage the replacement schema as a basis for a novel interpretable model for semantic change. We are also the first to evaluate the use of LLaMa for semantic change detection.

Recent advancements in large language models (LLMs) have significantly advanced the capabilities of summarization systems.However, they continue to face a persistent challenge: hallucination. While prior work has extensively examined LLMs in news domains, evaluation of dialogue summarization has primarily focused on BART-based models, resulting in a notable gap in understanding LLM effectiveness.Our work seeks to address this gap by benchmarking LLMs for dialogue summarization faithfulness using human annotations,focusing on identifying and categorizing span-level inconsistencies.Specifically, we evaluate two prominent LLMs: GPT-4 and Alpaca-13B.Our evaluation reveals that LLMs often generate plausible, but not fully supported inferences based on conversation contextual cues, a trait absent in older models. As a result, we propose a refined taxonomy of errors, introducing a novel category termed “Contextual Inference” to address this aspect of LLM behavior. Using our taxonomy, we compare the behavioral differences between LLMs and older fine-tuned models. Additionally, we systematically assess the efficacy of automatic error detection methods on LLM summaries and find that they struggle to detect these nuanced errors effectively. To address this, we introduce two prompt-based approaches for fine-grained error detection. Our methods outperform existing metrics, particularly in identifying the novel “Contextual Inference” error type.

Query expansion (QE) is a critical component in the open-domain question answering (OpenQA) pipeline, enhancing the retrieval performance by broadening the scope of queries with additional relevant texts. However, existing methods like GAR and EAR rely heavily on supervised training and often struggle to maintain effectiveness across domains and datasets. Meanwhile, although large language models (LLMs) have demonstrated QE capability for information retrieval (IR) tasks, their application in OpenQA is hindered by the inadequate analysis of query’s informational needs and the lack of quality control for generated QEs, failing to meet the unique requirements of OpenQA. To bridge this gap, we propose a novel LLM-based QE approach named AGR for the OpenQA task, leveraging a three-step prompting strategy. AGR begins with an analysis of the query, followed by the generation of answer-oriented expansions, and culminates with a refinement process for better query formulation. Extensive experiments on four OpenQA datasets reveal that AGR not only rivals in-domain supervised methods in retrieval accuracy, but also outperforms state-of-the-art baselines in out-domain zero-shot scenarios. Moreover, it exhibits enhanced performance in end-to-end QA evaluations, underscoring the superiority of AGR for OpenQA.