In this paper we provide evidence that a virtual model of U.S. congresspersons based on a collection of language models satisfies the definition of a digital twin. In particular, we introduce and provide high-level descriptions of a daily-updated dataset that contains every Tweet from every U.S. congressperson during their respective terms. We demonstrate that a modern language model equipped with congressperson-specific subsets of this data are capable of producing Tweets that are largely indistinguishable from actual Tweets posted by their physical counterparts. We illustrate how generated Tweets can be used to predict roll-call vote behaviors and to quantify the likelihood of congresspersons crossing party lines, thereby assisting stakeholders in allocating resources and potentially impacting real-world legislative dynamics. We conclude with a discussion of the limitations and important extensions of our analysis.
Understanding source code is a topic of great interest in the software engineering community, since it can help programmers in various tasks such as software maintenance and reuse. Recent advances in large language models (LLMs) have demonstrated remarkable program comprehension capabilities, while transformer-based topic modeling techniques offer effective ways to extract semantic information from text. This paper proposes and explores a novel approach that combines these strengths to automatically identify meaningful topics in a corpus of Python programs. Our method consists in applying topic modeling on the descriptions obtained by asking an LLM to summarize the code. To assess the internal consistency of the extracted topics, we compare them against topics inferred from function names alone, and those derived from existing docstrings. Experimental results suggest that leveraging LLM-generated summaries provides interpretable and semantically rich representation of code structure. The promising results suggest that our approach can be fruitfully applied in various software engineering tasks such as automatic documentation and tagging, code search, software reorganization and knowledge discovery in large repositories.
Large language models (LLMs) play a crucial role in software engineering, excelling in tasks like code generation and maintenance. However, existing benchmarks are often narrow in scope, focusing on a specific task and lack a comprehensive evaluation framework that reflects real-world applications. To address these gaps, we introduce CoCo-Bench (Comprehensive Code Benchmark), designed to evaluate LLMs across four critical dimensions: code understanding, code generation, code modification, and code review. These dimensions capture essential developer needs, ensuring a more systematic and representative evaluation. CoCo-Bench includes multiple programming languages and varying task difficulties, with rigorous manual review to ensure data quality and accuracy. Empirical results show that CoCo-Bench aligns with existing benchmarks while uncovering significant variations in model performance, effectively highlighting strengths and weaknesses. By offering a holistic and objective evaluation, CoCo-Bench provides valuable insights to guide future research and technological advancements in code-oriented LLMs, establishing a reliable benchmark for the field.
This project introduces BrAIcht, an AI conversational agent that creates dialogues in the distinctive style of the famous German playwright Bertolt Brecht. BrAIcht is fine-tuned using German LeoLM, a large language model with 7 billion parameters and a modified version of the base Llama2 suitable for German language tasks. For fine-tuning, 29 plays of Bertolt Brecht and 907 of other German plays that are stylistically similar to Bertolt Brecht are used to form a more di-erse dataset. Due to the limited memory capacity, a parameterefficient fine-tuning technique called QLoRA is implemented to train the large language model. The results, based on BLEU score and perplexity, show very promising performance of BrAIcht in generating dialogues in the style of Bertolt Brecht.
We present UniDetox, a universally applicable method designed to mitigate toxicity across various large language models (LLMs). Previous detoxification methods are typically model-specific, addressing only individual models or model families, and require careful hyperparameter tuning due to the trade-off between detoxification efficacy and language modeling performance. In contrast, UniDetox provides a detoxification technique that can be universally applied to a wide range of LLMs without the need for separate model-specific tuning. Specifically, we propose a novel and efficient dataset distillation technique for detoxification using contrastive decoding. This approach distills detoxifying representations in the form of synthetic text data, enabling universal detoxification of any LLM through fine-tuning with the distilled text. Our experiments demonstrate that the detoxifying text distilled from GPT-2 can effectively detoxify larger models, including OPT, Falcon, and LLaMA-2. Furthermore, UniDetox eliminates the need for separate hyperparameter tuning for each model, as a single hyperparameter configuration can be seamlessly applied across different models. Additionally, analysis of the detoxifying text reveals a reduction in politically biased content, providing insights into the attributes necessary for effective detoxification of LLMs.
In arbitrary-order language models, it is an open question how to sample tokens in parallel from the correct joint distribution. With discrete diffusion models, the more tokens they generate in parallel, the less their predicted distributions adhere to the originally learned data distribution, as they rely on a conditional independence assumption that only works with infinitesimally small timesteps. We find that a different class of models, any-subset autoregressive models (AS-ARMs), holds the solution. As implied by the name, AS-ARMs can generate tokens in any order, and in parallel. Moreover, AS-ARMs support parallelized joint probability density estimation, allowing them to correct their own parallel-generated token distributions, via our Any-Subset Speculative Decoding (ASSD) algorithm. ASSD provably enables generation of tokens from the correct joint distribution, with the number of neural network calls upper bounded by the number of tokens predicted. We empirically verify that ASSD speeds up language generation, without sacrificing quality. Furthermore, we provide a mathematically justified scheme for training AS-ARMs for generation, and show that AS-ARMs achieve state-of-the-art performance among sub-200M parameter models on infilling benchmark tasks, and nearly match the performance of models 50X larger on code generation. Our theoretical and empirical results indicate that the once-forgotten AS-ARMs are a promising direction of language modeling.
Large Language Models (LLMs) have demonstrated significant potential in algorithm design. However, when integrated into search frameworks for iterative algorithm search, the underlying fitness landscape--critical for understanding search behaviou--remains underexplored. In this paper, we illustrate and analyze the fitness landscape of LLM-assisted Algorithm Search (LAS) using a graph-based approach, where nodes represent algorithms and edges denote transitions between them. We conduct extensive evaluations across six algorithm design tasks and six commonly used LLMs. Our findings reveal that LAS landscapes are highly multimodal and rugged, particularly in combinatorial optimization tasks, with distinct structural variations across tasks and LLMs. For instance, heuristic design tasks exhibit dense clusters of high-performing algorithms, while symbolic regression tasks show sparse, scattered distributions. Additionally, we demonstrate how population size influences exploration-exploitation trade-offs and the evolving trajectory of elite algorithms. These insights not only advance our understanding of LAS landscapes but also provide practical guidance for designing more effective LAS methods.
The trend towards large language models (LLMs) for guardrailing against undesired behaviors is increasing and has shown promise for censoring user inputs. However, increased latency, memory consumption, hosting expenses and non-structured outputs can make their use prohibitive. In this work, we show that task-specific data generation can lead to fine-tuned classifiers that significantly outperform current state of the art (SoTA) while being orders of magnitude smaller. Secondly, we show that using a single model, \texttt{MultiTaskGuard}, that is pretrained on a large synthetically generated dataset with unique task instructions further improves generalization. Thirdly, our most performant models, \texttt{UniGuard}, are found using our proposed search-based model merging approach that finds an optimal set of parameters to combine single-policy models and multi-policy guardrail models. % On 7 public datasets and 4 guardrail benchmarks we created, our efficient guardrail classifiers improve over the best performing SoTA publicly available LLMs and 3$^{\text{rd}}$ party guardrail APIs in detecting unsafe and safe behaviors by an average F1 score improvement of \textbf{29.92} points over Aegis-LlamaGuard and \textbf{21.62} over \texttt{gpt-4o}, respectively. Lastly, our guardrail synthetic data generation process that uses custom task-specific guardrail poli
Soccer is a globally popular sporting event, typically characterized by long matches and distinctive highlight moments. Recent advances in Multimodal Large Language Models (MLLMs) offer promising capabilities in temporal grounding and video understanding, soccer commentary generation often requires precise temporal localization and semantically rich descriptions over long-form video. However, existing soccer MLLMs often rely on the temporal a priori for caption generation, so they cannot process the soccer video end-to-end. While some traditional approaches follow a two-step paradigm that is complex and fails to capture the global context to achieve suboptimal performance. To solve the above issues, we present TimeSoccer, the first end-to-end soccer MLLM for Single-anchor Dense Video Captioning (SDVC) in full-match soccer videos. TimeSoccer jointly predicts timestamps and generates captions in a single pass, enabling global context modeling across 45-minute matches. To support long video understanding of soccer matches, we introduce MoFA-Select, a training-free, motion-aware frame compression module that adaptively selects representative frames via a coarse-to-fine strategy, and incorporates complementary training paradigms to strengthen the model's ability to handle long temporal sequences. Extensive experiments demonstrate that our TimeSoccer achieves State-of-The-Art (SoTA) performance on the SDVC task in an end-to-end form, generating high-quality commentary with accurate temporal alignment and strong semantic relevance.
This paper introduces DATETIME, a new high-quality benchmark designed to evaluate the translation and reasoning abilities of a Large Language Model (LLM) on datetimes. A datetime is simply a date and a time, for example '11th.february.2023 ,1:12:31'. Datetimes are an interesting domain because they are intuitive and straightforward for humans to process but present significant challenges for LLMs. At the time of writing, no publicly available benchmark exists for systematically evaluating LLMs on datetime processing. Our experiments show that state-of-the-art models exhibit significant difficulty with tasks involving reasoning on datetimes, and that General Artificial Intelligence is still a distant aspiration. We hypothesize that working with datetimes necessitates translation and/or computation capabilities, and the tasks of the benchmark are organized accordingly. Significant dispersion in performance across models is observed with surprisingly poor performance even on apparently trivial tasks. Whilst frontier models such as ChatGPT, Claude and Llama3.1 have evidently been built and trained with datetime reasoning abilities, significant improvement is required for the open-source models.
We propose a novel spectral generative modeling framework for natural language processing that jointly learns a global time varying Fourier dictionary and per token mixing coefficients, replacing the ubiquitous self attention mechanism in transformer architectures. By enforcing reconstruction losses in both the time domain (embedding reconstruction) and the frequency domain (via Short Time Fourier Transform magnitude matching) alongside a standard language modeling objective, and fitting a Gaussian Mixture Model (GMM) prior over the learned mixing vectors, our approach achieves competitive perplexity and generation quality on standard benchmarks such as WikiText2 and Penn Treebank. In contrast to the quadratic computation complexity of self attention, our method operates with linear complexity, delivering substantial efficiency gains. We demonstrate that spectral dictionary models can achieve competitive performance compared to transformer baselines while significantly reducing inference latency and memory footprint, offering a compelling alternative for scalable language modeling.
Large language models (LLMs) exhibit probabilistic output characteristics, yet conventional evaluation frameworks rely on deterministic scalar metrics. This study introduces a Bayesian approach for LLM capability assessment that integrates prior knowledge through probabilistic inference, addressing limitations under limited-sample regimes. By treating model capabilities as latent variables and leveraging a curated query set to induce discriminative responses, we formalize model ranking as a Bayesian hypothesis testing problem over mutually exclusive capability intervals. Experimental evaluations with GPT-series models demonstrate that the proposed method achieves superior discrimination compared to conventional evaluation methods. Results indicate that even with reduced sample sizes, the approach maintains statistical robustness while providing actionable insights, such as probabilistic statements about a model's likelihood of surpassing specific baselines. This work advances LLM evaluation methodologies by bridging Bayesian inference with practical constraints in real-world deployment scenarios.
This study aims to guide language model selection by investigating: 1) the necessity of finetuning versus zero-shot usage, 2) the benefits of domain-adjacent versus generic pretrained models, 3) the value of further domain-specific pretraining, and 4) the continued relevance of Small Language Models (SLMs) compared to Large Language Models (LLMs) for specific tasks. Using electronic pathology reports from the British Columbia Cancer Registry (BCCR), three classification scenarios with varying difficulty and data size are evaluated. Models include various SLMs and an LLM. SLMs are evaluated both zero-shot and finetuned; the LLM is evaluated zero-shot only. Finetuning significantly improved SLM performance across all scenarios compared to their zero-shot results. The zero-shot LLM outperformed zero-shot SLMs but was consistently outperformed by finetuned SLMs. Domain-adjacent SLMs generally performed better than the generic SLM after finetuning, especially on harder tasks. Further domain-specific pretraining yielded modest gains on easier tasks but significant improvements on the complex, data-scarce task. The results highlight the critical role of finetuning for SLMs in specialized domains, enabling them to surpass zero-shot LLM performance on targeted classification tasks. Pretraining on domain-adjacent or domain-specific data provides further advantages, particularly for complex problems or limited finetuning data. While LLMs offer strong zero-shot capabilities, their performance on these specific tasks did not match that of appropriately finetuned SLMs. In the era of LLMs, SLMs remain relevant and effective, offering a potentially superior performance-resource trade-off compared to LLMs.
Large Language Models (LLMs), such as ChatGPT, have demonstrated the capability to generate human like, natural responses across a range of tasks, including task oriented dialogue and question answering. However, their application in real world, critical scenarios is often hindered by a tendency to produce inaccurate information and a limited ability to leverage external knowledge sources. This paper introduces the LLM ENHANCER system, designed to integrate multiple online sources such as Google, Wikipedia, and DuckDuckGo to enhance data accuracy. The LLMs employed within this system are open source. The data acquisition process for the LLM ENHANCER system operates in parallel, utilizing custom agent tools to manage the flow of information. Vector embeddings are used to identify the most pertinent information, which is subsequently supplied to the LLM for user interaction. The LLM ENHANCER system mitigates hallucinations in chat based LLMs while preserving response naturalness and accuracy.
This paper examines language modeling based on the theory of quantum mechanics. It focuses on the introduction of quantum mechanics into the symbol-meaning pairs of language in order to build a representation model of natural language. At the same time, it is realized that word embedding, which is widely used as a basic technique for statistical language modeling, can be explained and improved by the mathematical framework of quantum mechanics. On this basis, this paper continues to try to use quantum statistics and other related theories to study the mathematical representation, natural evolution and statistical properties of natural language. It is also assumed that the source of such quantum properties is the physicality of information. The feasibility of using quantum theory to model natural language is pointed out through the construction of a experimental code. The paper discusses, in terms of applications, the possible help of the theory in constructing generative models that are popular nowadays. A preliminary discussion of future applications of the theory to quantum computers is also presented.
With the rapid development and widespread application of Large Language Models (LLMs), multidimensional evaluation has become increasingly critical. However, current evaluations are often domain-specific and overly complex, limiting their effectiveness as cross-domain proxies for core capabilities. To address these limitations and enable a unified and simple evaluation framework, an ideal proxy task should target a basic capability that generalizes across tasks and is independent of domain-specific knowledge. Turing machine provides a powerful theoretical lens by reducing complex processes to basic, domain-agnostic computational operations. This perspective offers a principled framework for evaluating basic computational abilities essential to a wide range of tasks. Motivated by this abstraction, we introduce \textbf{Turing Machine Bench}, a benchmark designed to assess the ability of LLMs to \textbf{strictly follow rules} and \textbf{accurately manage internal states} for multi-step, referred to as \textbf{computational reasoning}. TMBench incorporates four key features: self-contained and knowledge-agnostic reasoning, a minimalistic multi-step structure, controllable difficulty, and a solid theoretical foundation based on Turing machine. Empirical results demonstrate that TMBench serves as an effective proxy for evaluating computational reasoning on representative LLMs. It produces clear step-wise accuracy curves, revealing LLMs' ability to execute multi-step reasoning processes. By analyzing performance trends across TMBench and established reasoning benchmarks, we find strong correlations with real-world tasks, bridging real-task evaluation with basic ability assessment. These findings suggest that TMBench holds potential as a cross-domain dimension for evaluating reasoning in LLMs. Code and data are available at \href{https://github.com/HaitaoWuTJU/Turing-Machine-Bench}{Repo}.
Classical Chinese, as the core carrier of Chinese culture, plays a crucial role in the inheritance and study of ancient literature. However, existing natural language processing models primarily optimize for Modern Chinese, resulting in inadequate performance on Classical Chinese. This paper presents a comprehensive solution for Classical Chinese language processing. By continuing pre-training and instruction fine-tuning on the LLaMA3-8B-Chinese model, we construct a large language model, WenyanGPT, which is specifically designed for Classical Chinese tasks. Additionally, we develop an evaluation benchmark dataset, WenyanBENCH. Experimental results on WenyanBENCH demonstrate that WenyanGPT significantly outperforms current advanced LLMs in various Classical Chinese tasks. We make the model's training data, instruction fine-tuning data\footnote, and evaluation benchmark dataset publicly available to promote further research and development in the field of Classical Chinese processing.
This paper introduces MoxE, a novel architecture that synergistically combines the Extended Long Short-Term Memory (xLSTM) with the Mixture of Experts (MoE) framework to address critical scalability and efficiency challenges in large language models (LLMs). The proposed method effectively leverages xLSTM's innovative memory structures while strategically introducing sparsity through MoE to substantially reduce computational overhead. At the heart of our approach is a novel entropy-based routing mechanism, designed to dynamically route tokens to specialized experts, thereby ensuring efficient and balanced resource utilization. This entropy awareness enables the architecture to effectively manage both rare and common tokens, with mLSTM blocks being favored to handle rare tokens. To further enhance generalization, we introduce a suite of auxiliary losses, including entropy-based and group-wise balancing losses, ensuring robust performance and efficient training. Theoretical analysis and empirical evaluations rigorously demonstrate that MoxE achieves significant efficiency gains and enhanced effectiveness compared to existing approaches, marking a notable advancement in scalable LLM architectures.
Simulation-based testing is crucial for validating autonomous vehicles (AVs), yet existing scenario generation methods either overfit to common driving patterns or operate in an offline, non-interactive manner that fails to expose rare, safety-critical corner cases. In this paper, we introduce an online, retrieval-augmented large language model (LLM) framework for generating safety-critical driving scenarios. Our method first employs an LLM-based behavior analyzer to infer the most dangerous intent of the background vehicle from the observed state, then queries additional LLM agents to synthesize feasible adversarial trajectories. To mitigate catastrophic forgetting and accelerate adaptation, we augment the framework with a dynamic memorization and retrieval bank of intent-planner pairs, automatically expanding its behavioral library when novel intents arise. Evaluations using the Waymo Open Motion Dataset demonstrate that our model reduces the mean minimum time-to-collision from 1.62 to 1.08 s and incurs a 75% collision rate, substantially outperforming baselines.
We present the design process and findings of the pre-conference workshop at the Machine Learning for Healthcare Conference (2024) entitled Red Teaming Large Language Models for Healthcare, which took place on August 15, 2024. Conference participants, comprising a mix of computational and clinical expertise, attempted to discover vulnerabilities -- realistic clinical prompts for which a large language model (LLM) outputs a response that could cause clinical harm. Red-teaming with clinicians enables the identification of LLM vulnerabilities that may not be recognised by LLM developers lacking clinical expertise. We report the vulnerabilities found, categorise them, and present the results of a replication study assessing the vulnerabilities across all LLMs provided.