Christoph Weisser

Anton Frederik Thielmann, Manish Kumar, Christoph Weisser et al.

The analysis of tabular data has traditionally been dominated by gradient-boosted decision trees (GBDTs), known for their proficiency with mixed categorical and numerical features. However, recent deep learning innovations are challenging this dominance. This paper investigates the use of autoregressive state-space models for tabular data and compares their performance against established benchmark models. Additionally, we explore various adaptations of these models, including different pooling strategies, feature interaction mechanisms, and bi-directional processing techniques to understand their effectiveness for tabular data. Our findings indicate that interpreting features as a sequence and processing them and their interactions through structured state-space layers can lead to significant performance improvement. This research underscores the versatility of autoregressive models in tabular data analysis, positioning them as a promising alternative that could substantially enhance deep learning capabilities in this traditionally challenging area. The source code is available at https://github.com/basf/mamba-tabular.

Andreas Buchmüller, Gillian Kant, Christoph Weisser et al.

We present Twitmo, a package that provides a broad range of methods to collect, pre-process, analyze and visualize geo-tagged Twitter data. Twitmo enables the user to collect geo-tagged Tweets from Twitter and and provides a comprehensive and user-friendly toolbox to generate topic distributions from Latent Dirichlet Allocations (LDA), correlated topic models (CTM) and structural topic models (STM). Functions are included for pre-processing of text, model building and prediction. In addition, one of the innovations of the package is the automatic pooling of Tweets into longer pseudo-documents using hashtags and cosine similarities for better topic coherence. The package additionally comes with functionality to visualize collected data sets and fitted models in static as well as interactive ways and offers built-in support for model visualizations via LDAvis providing great convenience for researchers in this area. The Twitmo package is an innovative toolbox that can be used to analyze public discourse of various topics, political parties or persons of interest in space and time.

Fabian Lukassen, Christoph Weisser, Michael Schlee et al.

This paper introduces a novel changepoint detection framework that combines ensemble statistical methods with Large Language Models (LLMs) to enhance both detection accuracy and the interpretability of regime changes in time series data. Two critical limitations in the field are addressed. First, individual detection methods exhibit complementary strengths and weaknesses depending on data characteristics, making method selection non-trivial and prone to suboptimal results. Second, automated, contextual explanations for detected changes are largely absent. The proposed ensemble method aggregates results from ten distinct changepoint detection algorithms, achieving superior performance and robustness compared to individual methods. Additionally, an LLM-powered explanation pipeline automatically generates contextual narratives, linking detected changepoints to potential real-world historical events. For private or domain-specific data, a Retrieval-Augmented Generation (RAG) solution enables explanations grounded in user-provided documents. The open source Python framework demonstrates practical utility in diverse domains, including finance, political science, and environmental science, transforming raw statistical output into actionable insights for analysts and decision-makers.

Fabian Lukassen, Jan Herrmann, Christoph Weisser et al.

Prior work shows that Large Language Models (LLMs) can transform Explainable AI (XAI) outputs into Natural Language Explanations (NLEs) that score highly on quality metrics such as plausibility, coherence, and comprehensibility. But does explanation quality translate to practical usefulness? We investigate this question in a time-series energy forecasting domain through five controlled experiments (2,730 judgments across 60 test instances), each operationalising a distinct facet of usefulness studied in the XAI literature. Holding NLE quality constant at the high levels established by a prior factorial study, we find that NLEs do not improve task accuracy on any of the five tasks, while inflating self-reported confidence. A placebic control shows that this confidence boost is driven by text presence rather than content. In an out-of-distribution detection task, NLEs reduce the LLM judge's ability to flag unreliable predictions, providing false reassurance that masks model failure. We characterise these findings as the Quality-Usefulness Gap and argue that evaluation of the XAI-to-NLE pipeline must extend beyond text-quality metrics to downstream task performance.

Parastoo Semnani, Mihail Bogojeski, Florian Bley et al.

The successful application of machine learning (ML) in catalyst design relies on high-quality and diverse data to ensure effective generalization to novel compositions, thereby aiding in catalyst discovery. However, due to complex interactions, catalyst design has long relied on trial-and-error, a costly and labor-intensive process leading to scarce data that is heavily biased towards undesired, low-yield catalysts. Despite the rise of ML in this field, most efforts have not focused on dealing with the challenges presented by such experimental data. To address these challenges, we introduce a robust machine learning and explainable AI (XAI) framework to accurately classify the catalytic yield of various compositions and identify the contributions of individual components. This framework combines a series of ML practices designed to handle the scarcity and imbalance of catalyst data. We apply the framework to classify the yield of various catalyst compositions in oxidative methane coupling, and use it to evaluate the performance of a range of ML models: tree-based models, logistic regression, support vector machines, and neural networks. These experiments demonstrate that the methods used in our framework lead to a significant improvement in the performance of all but one of the evaluated models. Additionally, the decision-making process of each ML model is analyzed by identifying the most important features for predicting catalyst performance using XAI methods. Our analysis found that XAI methods, providing class-aware explanations, such as Layer-wise Relevance Propagation, identified key components that contribute specifically to high-yield catalysts. These findings align with chemical intuition and existing literature, reinforcing their validity. We believe that such insights can assist chemists in the development and identification of novel catalysts with superior performance.

Anton Thielmann, Christoph Weisser, Benjamin Säfken

Few-shot methods for accurate modeling under sparse label-settings have improved significantly. However, the applications of few-shot modeling in natural language processing remain solely in the field of document classification. With recent performance improvements, supervised few-shot methods, combined with a simple topic extraction method pose a significant challenge to unsupervised topic modeling methods. Our research shows that supervised few-shot learning, combined with a simple topic extraction method, can outperform unsupervised topic modeling techniques in terms of generating coherent topics, even when only a few labeled documents per class are used.

Manish Kumar, Anton Frederik Thielmann, Christoph Weisser et al.

Numerical preprocessing remains an important component of tabular deep learning, where the representation of continuous features can strongly affect downstream performance. Although its importance is well established for classical statistical and machine learning models, the role of explicit numerical preprocessing in tabular deep learning remains less well understood. In this work, we study this question with a focus on spline-based numerical encodings. We investigate three spline families for encoding numerical features, namely B-splines, M-splines, and integrated splines (I-splines), under uniform, quantile-based, target-aware, and learnable-knot placement. For the learnable-knot variants, we use a differentiable knot parameterization that enables stable end-to-end optimization of knot locations jointly with the backbone. We evaluate these encodings on a diverse collection of public regression and classification datasets using MLP, ResNet, and FT-Transformer backbones, and compare them against common numerical preprocessing baselines. Our results show that the effect of numerical encodings depends strongly on the task, output size, and backbone. For classification, piecewise-linear encoding (PLE) is the most robust choice overall, while spline-based encodings remain competitive. For regression, no single encoding dominates uniformly. Instead, performance depends on the spline family, knot-placement strategy, and output size, with larger gains typically observed for MLP and ResNet than for FT-Transformer. We further find that learnable-knot variants can be optimized stably under the proposed parameterization, but may substantially increase training cost, especially for M-spline and I-spline expansions. Overall, the results show that numerical encodings should be assessed not only in terms of predictive performance, but also in terms of computational overhead.

Arik Reuter, Bishnu Khadka, Anton Thielmann et al.

Topic modeling seems to be almost synonymous with generating lists of top words to represent topics within large text corpora. However, deducing a topic from such list of individual terms can require substantial expertise and experience, making topic modelling less accessible to people unfamiliar with the particularities and pitfalls of top-word interpretation. A topic representation limited to top-words might further fall short of offering a comprehensive and easily accessible characterization of the various aspects, facets and nuances a topic might have. To address these challenges, we introduce GPTopic, a software package that leverages Large Language Models (LLMs) to create dynamic, interactive topic representations. GPTopic provides an intuitive chat interface for users to explore, analyze, and refine topics interactively, making topic modeling more accessible and comprehensive. The corresponding code is available here: https://github.com/ArikReuter/TopicGPT.

Arik Reuter, Anton Thielmann, Christoph Weisser et al.

Topic modelling was mostly dominated by Bayesian graphical models during the last decade. With the rise of transformers in Natural Language Processing, however, several successful models that rely on straightforward clustering approaches in transformer-based embedding spaces have emerged and consolidated the notion of topics as clusters of embedding vectors. We propose the Transformer-Representation Neural Topic Model (TNTM), which combines the benefits of topic representations in transformer-based embedding spaces and probabilistic modelling. Therefore, this approach unifies the powerful and versatile notion of topics based on transformer embeddings with fully probabilistic modelling, as in models such as Latent Dirichlet Allocation (LDA). We utilize the variational autoencoder (VAE) framework for improved inference speed and modelling flexibility. Experimental results show that our proposed model achieves results on par with various state-of-the-art approaches in terms of embedding coherence while maintaining almost perfect topic diversity. The corresponding source code is available at https://github.com/ArikReuter/TNTM.

Alvaro Paredes Amorin, Andre Python, Christoph Weisser

Large language models (LLMs) play an increasingly important role in finan- cial markets analysis by capturing signals from complex and heterogeneous textual data sources, such as tweets, news articles, reports, and microblogs. However, their performance is dependent on large computational resources and proprietary datasets, which are costly, restricted, and therefore inacces- sible to many researchers and practitioners. To reflect realistic situations we investigate the ability of lightweight open-source LLMs - smaller and publicly available models designed to operate with limited computational resources - to generalize sentiment understanding from financial datasets of varying sizes, sources, formats, and languages. We compare the benchmark finance natural language processing (NLP) model, FinBERT, and three open-source lightweight LLMs, DeepSeek-LLM 7B, Llama3 8B Instruct, and Qwen3 8B on five publicly available datasets: FinancialPhraseBank, Financial Question Answering, Gold News Sentiment, Twitter Sentiment and Chinese Finance Sentiment. We find that LLMs, specially Qwen3 8B and Llama3 8B, perform best in most scenarios, even from using only 5% of the available training data. These results hold in zero-shot and few-shot learning scenarios. Our findings indicate that lightweight, open-source large language models (LLMs) consti- tute a cost-effective option, as they can achieve competitive performance on heterogeneous textual data even when trained on only a limited subset of the extensive annotated corpora that are typically deemed necessary.

Fabian Lukassen, Jan Herrmann, Christoph Weisser et al.

Explainable AI (XAI) methods like SHAP and LIME produce numerical feature attributions that remain inaccessible to non expert users. Prior work has shown that Large Language Models (LLMs) can transform these outputs into natural language explanations (NLEs), but it remains unclear which factors contribute to high-quality explanations. We present a systematic factorial study investigating how Forecasting model choice, XAI method, LLM selection, and prompting strategy affect NLE quality. Our design spans four models (XGBoost (XGB), Random Forest (RF), Multilayer Perceptron (MLP), and SARIMAX - comparing black-box Machine-Learning (ML) against classical time-series approaches), three XAI conditions (SHAP, LIME, and a no-XAI baseline), three LLMs (GPT-4o, Llama-3-8B, DeepSeek-R1), and eight prompting strategies. Using G-Eval, an LLM-as-a-judge evaluation method, with dual LLM judges and four evaluation criteria, we evaluate 660 explanations for time-series forecasting. Our results suggest that: (1) XAI provides only small improvements over no-XAI baselines, and only for expert audiences; (2) LLM choice dominates all other factors, with DeepSeek-R1 outperforming GPT-4o and Llama-3; (3) we observe an interpretability paradox: in our setting, SARIMAX yielded lower NLE quality than ML models despite higher prediction accuracy; (4) zero-shot prompting is competitive with self-consistency at 7-times lower cost; and (5) chain-of-thought hurts rather than helps.

Alvaro Paredes Amorin, Andre Python, Christoph Weisser

By capturing the prevailing sentiment and market mood, textual data has become increasingly vital for forecasting commodity prices, particularly in metal markets. However, the effectiveness of lightweight, finetuned large language models (LLMs) in extracting predictive signals for aluminum prices, and the specific market conditions under which these signals are most informative, remains under-explored. This study generates monthly sentiment scores from English and Chinese news headlines (Reuters, Dow Jones Newswires, and China News Service) and integrates them with traditional tabular data, including base metal indices, exchange rates, inflation rates, and energy prices. We evaluate the predictive performance and economic utility of these models through long-short simulations on the Shanghai Metal Exchange from 2007 to 2024. Our results demonstrate that during periods of high volatility, Long Short-Term Memory (LSTM) models incorporating sentiment data from a finetuned Qwen3 model (Sharpe ratio 1.04) significantly outperform baseline models using tabular data alone (Sharpe ratio 0.23). Subsequent analysis elucidates the nuanced roles of news sources, topics, and event types in aluminum price forecasting.

Michael Schlee, Christoph Weisser, Timo Kivimäki et al.

LabelFusion is a fusion ensemble for text classification that learns to combine a traditional transformer-based classifier (e.g., RoBERTa) with one or more Large Language Models (LLMs such as OpenAI GPT, Google Gemini, or DeepSeek) to deliver accurate and cost-aware predictions across multi-class and multi-label tasks. The package provides a simple high-level interface (AutoFusionClassifier) that trains the full pipeline end-to-end with minimal configuration, and a flexible API for advanced users. Under the hood, LabelFusion integrates vector signals from both sources by concatenating the ML backbone's embeddings with the LLM-derived per-class scores -- obtained through structured prompt-engineering strategies -- and feeds this joint representation into a compact multi-layer perceptron (FusionMLP) that produces the final prediction. This learned fusion approach captures complementary strengths of LLM reasoning and traditional transformer-based classifiers, yielding robust performance across domains -- achieving 92.4% accuracy on AG News and 92.3% on 10-class Reuters 21578 topic classification -- while enabling practical trade-offs between accuracy, latency, and cost.

Mattias Luber, Anton Thielmann, Christoph Weisser et al.

Extracting topics from large collections of unstructured text-documents has become a central task in current NLP applications and algorithms like NMF, LDA as well as their generalizations are the well-established current state of the art. However, especially when it comes to short text documents like Tweets, these approaches often lead to unsatisfying results due to the sparsity of the document-feature matrices. Even though, several approaches have been proposed to overcome this sparsity by taking additional information into account, these are merely focused on the aggregation of similar documents and the estimation of word-co-occurrences. This ultimately completely neglects the fact that a lot of topical-information can be actually retrieved from so-called hashtag-graphs by applying common community detection algorithms. Therefore, this paper outlines a novel approach on how to integrate topic structures of hashtag graphs into the estimation of topic models by connecting graph-based community detection and semi-supervised NMF. By applying this approach on recently streamed Twitter data it will be seen that this procedure actually leads to more intuitive and humanly interpretable topics.