In this paper, we evaluate the creative fiction writing abilities of a fine-tuned small language model (SLM), BART-large, and compare its performance to human writers and two large language models (LLMs): GPT-3.5 and GPT-4o. Our evaluation consists of two experiments: (i) a human study in which 68 participants rated short stories from humans and the SLM on grammaticality, relevance, creativity, and attractiveness, and (ii) a qualitative linguistic analysis examining the textual characteristics of stories produced by each model. In the first experiment, BART-large outscored average human writers overall (2.11 vs. 1.85), a 14% relative improvement, though the slight human advantage in creativity was not statistically significant. In the second experiment, qualitative analysis showed that while GPT-4o demonstrated near-perfect coherence and used less cliche phrases, it tended to produce more predictable language, with only 3% of its synopses featuring surprising associations (compared to 15% for BART). These findings highlight how model size and fine-tuning influence the balance between creativity, fluency, and coherence in creative writing tasks, and demonstrate that smaller models can, in certain contexts, rival both humans and larger models.
With the advent of the big data and large language model era, zero-shot personalized rapid customization has emerged as a significant trend. In this report, we introduce Takin AudioLLM, a series of techniques and models, mainly including Takin TTS, Takin VC, and Takin Morphing, specifically designed for audiobook production. These models are capable of zero-shot speech production, generating high-quality speech that is nearly indistinguishable from real human speech and facilitating individuals to customize the speech content according to their own needs. Specifically, we first introduce Takin TTS, a neural codec language model that builds upon an enhanced neural speech codec and a multi-task training framework, capable of generating high-fidelity natural speech in a zero-shot way. For Takin VC, we advocate an effective content and timbre joint modeling approach to improve the speaker similarity, while advocating for a conditional flow matching based decoder to further enhance its naturalness and expressiveness. Last, we propose the Takin Morphing system with highly decoupled and advanced timbre and prosody modeling approaches, which enables individuals to customize speech production with their preferred timbre and prosody in a precise and controllable manner. Extensive experiments validate the effectiveness and robustness of our Takin AudioLLM series models. For detailed demos, please refer to https://everest-ai.github.io/takinaudiollm/.
Mixture-of-Experts (MoE) models scale more effectively than dense models due to sparse computation through expert routing, selectively activating only a small subset of expert modules. However, sparse computation challenges traditional training practices, as discrete expert routing hinders standard backpropagation and thus gradient-based optimization, which are the cornerstone of deep learning. To better pursue the scaling power of MoE, we introduce GRIN (GRadient-INformed MoE training), which incorporates sparse gradient estimation for expert routing and configures model parallelism to avoid token dropping. Applying GRIN to autoregressive language modeling, we develop a top-2 16$\times$3.8B MoE model. Our model, with only 6.6B activated parameters, outperforms a 7B dense model and matches the performance of a 14B dense model trained on the same data. Extensive evaluations across diverse tasks demonstrate the potential of GRIN to significantly enhance MoE efficacy, achieving 79.4 on MMLU, 83.7 on HellaSwag, 74.4 on HumanEval, and 58.9 on MATH.
Large Language Model can reasonably understand and generate human expressions but may lack of thorough thinking and reasoning mechanisms. Recently there have been several studies which enhance the thinking ability of language models but most of them are not data-driven or training-based. In this paper, we are motivated by the cognitive mechanism in the natural world, and design a novel model architecture called TaS which allows it to first consider the thoughts and then express the response based upon the query. We design several pipelines to annotate or generate the thought contents from prompt-response samples, then add language heads in a middle layer which behaves as the thinking layer. We train the language model by the thoughts-augmented data and successfully let the thinking layer automatically generate reasonable thoughts and finally output more reasonable responses. Both qualitative examples and quantitative results validate the effectiveness and performance of TaS. Our code is available at https://anonymous.4open.science/r/TadE.
LLM-as-a-Judge and reward models are widely used alternatives of multiple-choice questions or human annotators for large language model (LLM) evaluation. Their efficacy shines in evaluating long-form responses, serving a critical role as evaluators of leaderboards and as proxies to align LLMs via reinforcement learning. However, despite their popularity, their effectiveness in diverse contexts, such as non-English prompts, factual verification, or challenging questions, remains unexplored. In this paper, we conduct a comprehensive analysis of automated evaluators, reporting several key findings on their behavior. First, we discover that English evaluation capabilities significantly influence language-specific evaluation capabilities, often more than the language proficiency itself, enabling evaluators trained in English to easily transfer their skills to other languages. Second, we identify critical shortcomings, where LLMs fail to detect and penalize errors, such as factual inaccuracies, cultural misrepresentations, and the presence of unwanted language. Finally, we find that state-of-the-art evaluators struggle with challenging prompts, in either English or Korean, underscoring their limitations in assessing or generating complex reasoning questions. We release the dataset and codes used.
Aspect-based sentiment analysis (ABSA) involves identifying sentiment towards specific aspect terms in a sentence and allows us to uncover nuanced perspectives and attitudes on particular aspects of a product, service, or topic. However, the scarcity of labeled data poses a significant challenge to training high-quality models. To address this issue, we explore the potential of data augmentation using ChatGPT, a well-performing large language model (LLM), to enhance the sentiment classification performance towards aspect terms. Specifically, we explore three data augmentation strategies based on ChatGPT: context-focused, aspect-focused, and context-aspect data augmentation techniques. Context-focused data augmentation focuses on changing the word expression of context words in the sentence while keeping aspect terms unchanged. In contrast, aspect-focused data augmentation aims to change aspect terms but keep context words unchanged. Context-Aspect data augmentation integrates the above two data augmentations to generate augmented samples. Furthermore, we incorporate contrastive learning into the ABSA tasks to improve performance. Extensive experiments show that all three data augmentation techniques lead to performance improvements, with the context-aspect data augmentation strategy performing best and surpassing the performance of the baseline models.
Next-token prediction serves as the dominant component in current neural language models. During the training phase, the model employs teacher forcing, which predicts tokens based on all preceding ground truth tokens. However, this approach has been found to create shortcuts, utilizing the revealed prefix to spuriously fit future tokens, potentially compromising the accuracy of the next-token predictor. In this paper, we introduce Semformer, a novel method of training a Transformer language model that explicitly models the semantic planning of response. Specifically, we incorporate a sequence of planning tokens into the prefix, guiding the planning token representations to predict the latent semantic representations of the response, which are induced by an autoencoder. In a minimal planning task (i.e., graph path-finding), our model exhibits near-perfect performance and effectively mitigates shortcut learning, a feat that standard training methods and baseline models have been unable to accomplish. Furthermore, we pretrain Semformer from scratch with 125M parameters, demonstrating its efficacy through measures of perplexity, in-context learning, and fine-tuning on summarization tasks.
At the beginning of 2022, a simplistic word-guessing game took the world by storm and was further adapted to many languages beyond the original English version. In this paper, we examine the strategies of daily word-guessing game players that have evolved during a period of over two years. A survey gathered from 25% of frequent players reveals their strategies and motivations for continuing the daily journey. We also explore the capability of several popular open-access large language model systems and open-source models at comprehending and playing the game in two different languages. Results highlight the struggles of certain models to maintain correct guess length and generate repetitions, as well as hallucinations of non-existent words and inflections.
This paper proposes Attention-Seeker, an unsupervised keyphrase extraction method that leverages self-attention maps from a Large Language Model to estimate the importance of candidate phrases. Our approach identifies specific components - such as layers, heads, and attention vectors - where the model pays significant attention to the key topics of the text. The attention weights provided by these components are then used to score the candidate phrases. Unlike previous models that require manual tuning of parameters (e.g., selection of heads, prompts, hyperparameters), Attention-Seeker dynamically adapts to the input text without any manual adjustments, enhancing its practical applicability. We evaluate Attention-Seeker on four publicly available datasets: Inspec, SemEval2010, SemEval2017, and Krapivin. Our results demonstrate that, even without parameter tuning, Attention-Seeker outperforms most baseline models, achieving state-of-the-art performance on three out of four datasets, particularly excelling in extracting keyphrases from long documents.
Focalization, the perspective through which narrative is presented, is encoded via a wide range of lexico-grammatical features and is subject to reader interpretation. Even trained annotators frequently disagree on correct labels, suggesting this task is both qualitatively and computationally challenging. In this work, we test how well five contemporary large language model (LLM) families and two baselines perform when annotating short literary excerpts for focalization. Despite the challenging nature of the task, we find that LLMs show comparable performance to trained human annotators, with GPT-4o achieving an average F1 of 84.79%. Further, we demonstrate that the log probabilities output by GPT-family models frequently reflect the difficulty of annotating particular excerpts. Finally, we provide a case study analyzing sixteen Stephen King novels, demonstrating the usefulness of this approach for computational literary studies and the insights gleaned from examining focalization at scale.
This paper introduces Bio-Inspired Mamba (BIM), a novel online learning framework for selective state space models that integrates biological learning principles with the Mamba architecture. BIM combines Real-Time Recurrent Learning (RTRL) with Spike-Timing-Dependent Plasticity (STDP)-like local learning rules, addressing the challenges of temporal locality and biological plausibility in training spiking neural networks. Our approach leverages the inherent connection between backpropagation through time and STDP, offering a computationally efficient alternative that maintains the ability to capture long-range dependencies. We evaluate BIM on language modeling, speech recognition, and biomedical signal analysis tasks, demonstrating competitive performance against traditional methods while adhering to biological learning principles. Results show improved energy efficiency and potential for neuromorphic hardware implementation. BIM not only advances the field of biologically plausible machine learning but also provides insights into the mechanisms of temporal information processing in biological neural networks.
As the complexity and connectivity of networks increase, the need for novel malware detection approaches becomes imperative. Traditional security defenses are becoming less effective against the advanced tactics of today's cyberattacks. Deep Packet Inspection (DPI) has emerged as a key technology in strengthening network security, offering detailed analysis of network traffic that goes beyond simple metadata analysis. DPI examines not only the packet headers but also the payload content within, offering a thorough insight into the data traversing the network. This study proposes a novel approach that leverages a large language model (LLM) and few-shot learning to accurately recognizes novel, unseen malware types with few labels samples. Our proposed approach uses a pretrained LLM on known malware types to extract the embeddings from packets. The embeddings are then used alongside few labeled samples of an unseen malware type. This technique is designed to acclimate the model to different malware representations, further enabling it to generate robust embeddings for each trained and unseen classes. Following the extraction of embeddings from the LLM, few-shot learning is utilized to enhance performance with minimal labeled data. Our evaluation, which utilized two renowned datasets, focused on identifying malware types within network traffic and Internet of Things (IoT) environments. Our approach shows promising results with an average accuracy of 86.35% and F1-Score of 86.40% on different malware types across the two datasets.
This paper presents LOLA, a massively multilingual large language model trained on more than 160 languages using a sparse Mixture-of-Experts Transformer architecture. Our architectural and implementation choices address the challenge of harnessing linguistic diversity while maintaining efficiency and avoiding the common pitfalls of multilinguality. Our analysis of the evaluation results shows competitive performance in natural language generation and understanding tasks. Additionally, we demonstrate how the learned expert-routing mechanism exploits implicit phylogenetic linguistic patterns to potentially alleviate the curse of multilinguality. We provide an in-depth look at the training process, an analysis of the datasets, and a balanced exploration of the model's strengths and limitations. As an open-source model, LOLA promotes reproducibility and serves as a robust foundation for future research. Our findings enable the development of compute-efficient multilingual models with strong, scalable performance across languages.
Despite the impressive performance of autoregressive Language Models (LM) it has been shown that due to reporting bias, LMs lack visual knowledge, i.e. they do not know much about the visual world and its properties. To augment LMs with visual knowledge, existing solutions often rely on explicit images, requiring time-consuming retrieval or image generation systems. This paper shows that explicit images are not necessary to visually augment an LM. Instead, we use visually-grounded text representations obtained from the well-known CLIP multimodal system. For a fair comparison, we modify VALM, a visually-augmented LM which uses image retrieval and representation, to work directly with visually-grounded text representations. We name this new model BLIND-VALM. We show that BLIND-VALM performs on par with VALM for Visual Language Understanding (VLU), Natural Language Understanding (NLU) and Language Modeling tasks, despite being significantly more efficient and simpler. We also show that scaling up our model within the compute budget of VALM, either increasing the model or pre-training corpus size, we outperform VALM for all the evaluation tasks.
Exploring the narratives conveyed by fine-art paintings is a challenge in image captioning, where the goal is to generate descriptions that not only precisely represent the visual content but also offer a in-depth interpretation of the artwork's meaning. The task is particularly complex for artwork images due to their diverse interpretations and varied aesthetic principles across different artistic schools and styles. In response to this, we present KALE Knowledge-Augmented vision-Language model for artwork Elaborations), a novel approach that enhances existing vision-language models by integrating artwork metadata as additional knowledge. KALE incorporates the metadata in two ways: firstly as direct textual input, and secondly through a multimodal heterogeneous knowledge graph. To optimize the learning of graph representations, we introduce a new cross-modal alignment loss that maximizes the similarity between the image and its corresponding metadata. Experimental results demonstrate that KALE achieves strong performance (when evaluated with CIDEr, in particular) over existing state-of-the-art work across several artwork datasets. Source code of the project is available at https://github.com/Yanbei-Jiang/Artwork-Interpretation.
excel in delivering comprehensive suggestions by deeply analyzing content and user behavior. However, they often inherit biases from skewed training data, favoring mainstream content while underrepresenting diverse or non-traditional options. This study explores the interplay between bias and LLM-based recommendation systems, focusing on music, song, and book recommendations across diverse demographic and cultural groups. This paper analyzes bias in LLM-based recommendation systems across multiple models (GPT, LLaMA, and Gemini), revealing its deep and pervasive impact on outcomes. Intersecting identities and contextual factors, like socioeconomic status, further amplify biases, complicating fair recommendations across diverse groups. Our findings reveal that bias in these systems is deeply ingrained, yet even simple interventions like prompt engineering can significantly reduce it. We further propose a retrieval-augmented generation strategy to mitigate bias more effectively. Numerical experiments validate these strategies, demonstrating both the pervasive nature of bias and the impact of the proposed solutions.
Recent advancements in deep learning, particularly large language models (LLMs), made a significant impact on how researchers study microbiome and metagenomics data. Microbial protein and genomic sequences, like natural languages, form a language of life, enabling the adoption of LLMs to extract useful insights from complex microbial ecologies. In this paper, we review applications of deep learning and language models in analyzing microbiome and metagenomics data. We focus on problem formulations, necessary datasets, and the integration of language modeling techniques. We provide an extensive overview of protein/genomic language modeling and their contributions to microbiome studies. We also discuss applications such as novel viromics language modeling, biosynthetic gene cluster prediction, and knowledge integration for metagenomics studies.
Most instance perception and image understanding solutions focus mainly on natural images. However, applications for synthetic images, and more specifically, images of Graphical User Interfaces (GUI) remain limited. This hinders the development of autonomous computer-vision-powered Artificial Intelligence (AI) agents. In this work, we present Instruction Visual Grounding or IVG, a multi-modal solution for object identification in a GUI. More precisely, given a natural language instruction and GUI screen, IVG locates the coordinates of the element on the screen where the instruction would be executed. To this end, we develop two methods. The first method is a three-part architecture that relies on a combination of a Large Language Model (LLM) and an object detection model. The second approach uses a multi-modal foundation model.
Multimodal large language models (MLLMs) have achieved significant success in the general field of image processing. Their emerging task generalization and freeform conversational capabilities can greatly facilitate medical diagnostic assistance, helping patients better understand their conditions and enhancing doctor-patient trust. Computed Tomography (CT) is a non-invasive imaging technique used to capture the internal mechanisms of a patient's condition and is widely utilized. However, in past research, the complex textural features of this imaging data have made accurate interpretation by algorithms challenging, impeding the performance of general LLMs in diagnostic assistance. To address this, we developed OrthoDoc, a MLLM designed for CT diagnostics. OrthoDoc is trained on 120,000 CT images and diagnostic reports and includes a Retrieval-Augmented Generation (RAG) module capable of effectively mitigating model hallucinations. This module is informed by extensive medical literature, textbooks, and explanatory data. Thus, OrthoDoc not only processes complex CT images but also stores, understands, and reasons over medical knowledge and language. In extensive experiments, OrthoDoc outperforms commercial models led by GPT-4, demonstrating superior diagnostic capabilities and accuracy. Specifically, OrthoDoc significantly surpasses existing models in the diagnosis of common orthopedic conditions such as fractures, arthritis, and tumors. Additionally, OrthoDoc exhibits robust generalization and stability when handling rare and complex cases.
Large Language Models (LLMs), capable of handling multi-modal input and outputs such as text, voice, images, and video, are transforming the way we process information. Beyond just generating textual responses to prompts, they can integrate with different software platforms to offer comprehensive solutions across diverse applications. In this paper, we present ChatSUMO, a LLM-based agent that integrates language processing skills to generate abstract and real-world simulation scenarios in the widely-used traffic simulator - Simulation of Urban MObility (SUMO). Our methodology begins by leveraging the LLM for user input which converts to relevant keywords needed to run python scripts. These scripts are designed to convert specified regions into coordinates, fetch data from OpenStreetMap, transform it into a road network, and subsequently run SUMO simulations with the designated traffic conditions. The outputs of the simulations are then interpreted by the LLM resulting in informative comparisons and summaries. Users can continue the interaction and generate a variety of customized scenarios without prior traffic simulation expertise. For simulation generation, we created a real-world simulation for the city of Albany with an accuracy of 96\%. ChatSUMO also realizes the customizing of edge edit, traffic light optimization, and vehicle edit by users effectively.