Author name cluster

Thomas Hartvigsen

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

12 papers

2 author rows

TMLR Journal 2025 Journal Article

ADMIRE-BayesOpt: Accelerated Data MIxture RE-weighting for Language Models with Bayesian Optimization

Xu Ouyang
Shengzhuang Chen
Michael Arthur Leopold Pearce
Thomas Hartvigsen
Jonathan Richard Schwarz

Determining the optimal data mixture for large language model training remains a challenging problem with an outsized impact on performance. In practice, language model developers continue to rely on heuristic exploration since no learning-based approach has emerged as a reliable solution. In this work, we propose to view the selection of training data mixtures as a black-box hyperparameter optimization problem, for which Bayesian Optimization is a well-established class of appropriate algorithms. Firstly, we cast data mixture learning as a sequential decision-making problem, in which we aim to find a suitable trade-off between the computational cost of training exploratory (proxy-) models and final mixture performance. Secondly, we systematically explore the properties of transferring mixtures learned at a small scale to larger-scale experiments, providing insights and highlighting opportunities for research at a modest scale. By proposing Multi-fidelity Bayesian Optimization as a suitable method in this common scenario, we introduce a natural framework to balance experiment cost with model fit, avoiding the risks of overfitting to smaller scales while minimizing the number of experiments at high cost. We present results for pre-training and instruction finetuning across models ranging from 1 million to 7 billion parameters, varying from simple architectures to state-of-the-art models and benchmarks spanning dozens of datasets. We demonstrate consistently strong results relative to a wide range of benchmarks, showing a speed-ups of over 500% in determining the best data mixture on our largest experiments relative to recent baselines. In addition, we broaden access to research by sharing ADMIRE IFT Runs, a dataset of 460 full training & evaluation runs reproducible post-training pipelines worth over 13,000 GPU hours, greatly reducing the cost of conducting research in this area. Finally, we highlight rich opportunities for future research in this area, helping bridge the gap towards a comprehensive understanding of the broader effects of training data on model generalization.

PDF Details

ICML Conference 2025 Conference Paper

BalancEdit: Dynamically Balancing the Generality-Locality Trade-off in Multi-modal Model Editing

Dongliang Guo 0002
Mengxuan Hu
Zihan Guan 0001
Thomas Hartvigsen
Sheng Li 0001

Large multi-modal models inevitably decay over time as facts update and previously learned information becomes outdated. Traditional approaches such as fine-tuning are often impractical for updating these models due to their size and complexity. Instead, direct knowledge editing within the models presents a more viable solution. Current model editing techniques, however, typically overlook the unique influence ranges of different facts, leading to compromised model performance in terms of both generality and locality. To address this issue, we introduce the concept of the generality-locality trade-off in multi-modal model editing. We develop a new model editing dataset named OKEDIT, specifically designed to effectively evaluate this trade-off. Building on this foundation, we propose BalancEdit, a novel method for balanced model editing that dynamically achieves an optimal balance between generality and locality. BalancEdit utilizes a unique mechanism that generates both positive and negative samples for each fact to accurately determine its influence scope and incorporates these insights into the model’s latent space using a discrete, localized codebook of edits, without modifying the underlying model weights. To our knowledge, this is the first approach explicitly addressing the generality-locality trade-off in multi-modal model editing. Our comprehensive results confirm the effectiveness of BalancEdit, demonstrating minimal trade-offs while maintaining robust editing capabilities. Our code and dataset are available at https: //github. com/donglgcn/BalancEdit/tree/MMOKVQA.

Details

ICLR Conference 2025 Conference Paper

Composable Interventions for Language Models

Arinbjörn Kolbeinsson
Kyle O'Brien
Tianjin Huang
Shanghua Gao
Shiwei Liu 0003
Jonathan Richard Schwarz
Anurag J. Vaidya
Faisal Mahmood 0001

Test-time interventions for language models can enhance factual accuracy, mitigate harmful outputs, and improve model efficiency without costly retraining. But despite a flood of new methods, different types of interventions are largely developing independently. In practice, multiple interventions must be applied sequentially to the same model, yet we lack standardized ways to study how interventions interact. We fill this gap by introducing composable interventions, a framework to study the effects of using multiple interventions on the same language models, featuring new metrics and a unified codebase. Using our framework, we conduct extensive experiments and compose popular methods from three emerging intervention categories---knowledge editing, model compression, and machine unlearning. Our results over 417 different compositions uncover meaningful interactions: compression hinders editing and unlearning, composing interventions hinges on their order of application, and popular general-purpose metrics are inadequate for assessing composability. Taken together, our findings showcase clear gaps in composability, suggesting a need for new multi-objective interventions.

Details

ICLR Conference 2025 Conference Paper

Learning under Temporal Label Noise

Sujay Nagaraj
Walter Gerych
Sana Tonekaboni
Anna Goldenberg
Berk Ustun
Thomas Hartvigsen

Many time series classification tasks, where labels vary over time, are affected by label noise that also varies over time. Such noise can cause label quality to improve, worsen, or periodically change over time. We first propose and formalize temporal label noise, an unstudied problem for sequential classification of time series. In this setting, multiple labels are recorded over time while being corrupted by a time-dependent noise function. We first demonstrate the importance of modeling the temporal nature of the label noise function and how existing methods will consistently underperform. We then propose methods to train noise-tolerant classifiers by estimating the temporal label noise function directly from data. We show that our methods lead to state-of-the-art performance under diverse types of temporal label noise on real-world datasets.

Details

ICML Conference 2025 Conference Paper

WikiBigEdit: Understanding the Limits of Lifelong Knowledge Editing in LLMs

Lukas Thede
Karsten Roth
Matthias Bethge
Zeynep Akata
Thomas Hartvigsen

Keeping large language models factually up-to-date is crucial for deployment, yet costly retraining remains a challenge. Knowledge editing offers a promising alternative, but methods are only tested on small-scale or synthetic edit benchmarks. In this work, we aim to bridge research into lifelong knowledge editing to real-world edits at practically relevant scale. We first introduce WikiBigEdit; a large-scale benchmark of real-world Wikidata edits, built to automatically extend lifelong for future-proof benchmarking. In its first instance, it includes over 500K question-answer pairs for knowledge editing alongside a comprehensive evaluation pipeline. Finally, we use WikiBigEdit to study existing knowledge editing techniques’ ability to incorporate large volumes of real-world facts and contrast their capabilities to generic modification techniques such as retrieval augmentation and continual finetuning to acquire a complete picture of the practical extent of current lifelong knowledge editing.

Details

NeurIPS Conference 2024 Conference Paper

Are Language Models Actually Useful for Time Series Forecasting?

Mingtian Tan
Mike A. Merrill
Vinayak Gupta
Tim Althoff
Thomas Hartvigsen

Large language models (LLMs) are being applied to time series forecasting. But are language models actually useful for time series? In a series of ablation studies on three recent and popular LLM-based time series forecasting methods, we find that removing the LLM component or replacing it with a basic attention layer does not degrade forecasting performance---in most cases, the results even improve! We also find that despite their significant computational cost, pretrained LLMs do no better than models trained from scratch, do not represent the sequential dependencies in time series, and do not assist in few-shot settings. Additionally, we explore time series encoders and find that patching and attention structures perform similarly to LLM-based forecasters. All resources needed to reproduce our work are available: https: //github. com/BennyTMT/LLMsForTimeSeries.

PDF Details DOI

NeurIPS Conference 2024 Conference Paper

BendVLM: Test-Time Debiasing of Vision-Language Embeddings

Walter Gerych
Haoran Zhang
Kimia Hamidieh
Eileen Pan
Maanas Sharma
Thomas Hartvigsen
Marzyeh Ghassemi

Vision-language (VL) embedding models have been shown to encode biases present in their training data, such as societal biases that prescribe negative characteristics to members of various racial and gender identities. Due to their wide-spread adoption for various tasks ranging from few-shot classification to text-guided image generation, debiasing VL models is crucial. Debiasing approaches that fine-tune the VL model often suffer from catastrophic forgetting. On the other hand, fine-tuning-free methods typically utilize a ``one-size-fits-all" approach that assumes that correlation with the spurious attribute can be explained using a single linear direction across all possible inputs. In this work, we propose a nonlinear, fine-tuning-free approach for VL embedding model debiasing that tailors the debiasing operation to each unique input. This allows for a more flexible debiasing approach. Additionally, we do not require knowledge of the set of inputs a priori to inference time, making our method more appropriate for online tasks such as retrieval and text guided image generation.

PDF Details DOI

TMLR Journal 2024 Journal Article

Improving Black-box Robustness with In-Context Rewriting

Kyle O'Brien
Nathan Hoyen Ng
Isha Puri
Jorge Mendez-Mendez
Hamid Palangi
Yoon Kim
Marzyeh Ghassemi
Thomas Hartvigsen

Machine learning models for text classification often excel on in-distribution (ID) data but struggle with unseen out-of-distribution (OOD) inputs. Most techniques for improving OOD robustness are not applicable to settings where the model is effectively a black box, such as when the weights are frozen, retraining is costly, or the model is leveraged via an API. Test-time augmentation (TTA) is a simple post-hoc technique for improving robustness that sidesteps black-box constraints by aggregating predictions across multiple augmentations of the test input. TTA has seen limited use in NLP due to the challenge of generating effective natural language augmentations. In this work, we propose LLM-TTA, which uses LLM-generated augmentations as TTA's augmentation function. LLM-TTA outperforms conventional augmentation functions across sentiment, toxicity, and news classification tasks for BERT and T5 models, with BERT's OOD robustness improving by an average of 4.48 percentage points without regressing average ID performance. We explore selectively augmenting inputs based on prediction entropy to reduce the rate of expensive LLM augmentations, allowing us to maintain performance gains while reducing the average number of generated augmentations by 57.74\%. LLM-TTA is agnostic to the task model architecture, does not require OOD labels, and is effective across low and high-resource settings.

PDF Details

NeurIPS Conference 2024 Conference Paper

UniTS: A Unified Multi-Task Time Series Model

Shanghua Gao
Teddy Koker
Owen Queen
Thomas Hartvigsen
Theodoros Tsiligkaridis
Marinka Zitnik

Although pre-trained transformers and reprogrammed text-based LLMs have shown strong performance on time series tasks, the best-performing architectures vary widely across tasks, with most models narrowly focused on specific areas, such as time series forecasting. Unifying predictive and generative time series tasks within a single model remains challenging. We introduce UniTS, a unified multi-task time series model that utilizes task tokenization to integrate predictive and generative tasks into a single framework. UniTS employs a modified transformer block to capture universal time series representations, enabling transferability from a heterogeneous, multi-domain pre-training dataset—characterized by diverse dynamic patterns, sampling rates, and temporal scales—to a wide range of downstream datasets with varied task specifications and data domains. Tested on 38 datasets across human activity sensors, healthcare, engineering, and finance, UniTS achieves superior performance compared to 12 forecasting models, 20 classification models, 18 anomaly detection models, and 16 imputation models, including adapted text-based LLMs. UniTS also demonstrates strong few-shot and prompt capabilities when applied to new domains and tasks. In single-task settings, UniTS outperforms competitive task-specialized time series models. Code and datasets are available at https: //github. com/mims-harvard/UniTS.

PDF Details DOI

AAAI Conference 2023 Conference Paper

Knowledge Amalgamation for Multi-Label Classification via Label Dependency Transfer

Jidapa Thadajarassiri
Thomas Hartvigsen
Walter Gerych
Xiangnan Kong
Elke Rundensteiner

Multi-label classification (MLC), which assigns multiple labels to each instance, is crucial to domains from computer vision to text mining. Conventional methods for MLC require huge amounts of labeled data to capture complex dependencies between labels. However, such labeled datasets are expensive, or even impossible, to acquire. Worse yet, these pre-trained MLC models can only be used for the particular label set covered in the training data. Despite this severe limitation, few methods exist for expanding the set of labels predicted by pre-trained models. Instead, we acquire vast amounts of new labeled data and retrain a new model from scratch. Here, we propose combining the knowledge from multiple pre-trained models (teachers) to train a new student model that covers the union of the labels predicted by this set of teachers. This student supports a broader label set than any one of its teachers without using labeled data. We call this new problem knowledge amalgamation for multi-label classification. Our new method, Adaptive KNowledge Transfer (ANT), trains a student by learning from each teacher’s partial knowledge of label dependencies to infer the global dependencies between all labels across the teachers. We show that ANT succeeds in unifying label dependencies among teachers, outperforming five state-of-the-art methods on eight real-world datasets.

PDF Details DOI

AAAI Conference 2022 Conference Paper

Recovering the Propensity Score from Biased Positive Unlabeled Data

Walter Gerych
Thomas Hartvigsen
Luke Buquicchio
Emmanuel Agu
Elke Rundensteiner

Positive-Unlabeled (PU) learning methods train a classifier to distinguish between the positive and negative classes given only positive and unlabeled data. While traditional PU methods require the labeled positive samples to be an unbiased sample of the positive distribution, in practice the labeled sample is often a biased draw from the true distribution. Prior work shows that if we know the likelihood that each positive instance will be selected for labeling, referred to as the propensity score, then the biased sample can be used for PU learning. Unfortunately, no prior work has been proposed an inference strategy for which the propensity score is identifiable. In this work, we propose two sets of assumptions under which the propensity score can be uniquely determined: one in which no assumption is made on the functional form of the propensity score (requiring assumptions on the data distribution), and the second which loosens the data assumptions while assuming a functional form for the propensity score. We then propose inference strategies for each case. Our empirical study shows that our approach significantly outperforms the state-of-the-art propensity estimation methods on a rich variety of benchmark datasets.

PDF Details

AAAI Conference 2021 Conference Paper

Semi-Supervised Knowledge Amalgamation for Sequence Classification

Jidapa Thadajarassiri
Thomas Hartvigsen
Xiangnan Kong
Elke A Rundensteiner

Sequence classification is essential for domains from medical diagnosis to online advertising. In these settings, data are typically proprietary and annotations are expensive to acquire. Often times, so few annotations are available that training a robust model from scratch is impractical. Recently, knowledge amalgamation (KA) has emerged as a promising strategy for training models without this hard-to-come-by labeled training dataset. To achieve this, KA methods combine the knowledge of multiple pre-trained teacher models (trained on different classification tasks and proprietary datasets) into one student model that becomes an expert on the union of all teachers’ classes. However, we demonstrate that the state-ofthe-art solutions fail in the presence of overconfident teachers, which make confident but incorrect predictions for instances from classes upon which they were not trained. Additionally, to-date no work has explored KA for sequence models. Therefore, we propose and then solve the open problem of semi-supervised KA for sequence classification (SKA). Our SKA approach first learns to estimate how trustworthy each teacher is for a given instance, then rescales the predicted probabilities from all teachers to supervise a student model. Our solution overcomes overconfident teachers through careful use of a very small amount of labeled instances. We demonstrate that this approach beats eight state-of-the-art alternatives on four real-world datasets by on average 15% in accuracy with as little as 2% of training data being annotated.

PDF Details