10 papers · 1 filter
SNAP-KG: Streaming Node Assignment via Projection for Knowledge Graph Entity Integration
Jui-Chien Lin, Mohammad Mohammadi Amiri, Oshani Seneviratne
Knowledge graph (KG) construction pipelines must continuously integrate newly arriving entities into a growing graph. Unlike inserting triples between existing nodes, a newly arriv…
Curated Synthetic Data Doesn't Have to Collapse: A Theoretical Study of Generative Retraining with Pluralistic Preferences
Ali Falahati, Mohammad Mohammadi Amiri, Kate Larson +1
Recursive retraining of generative models poses a critical representation challenge: when synthetic outputs are curated based on a fixed reward signal, the model tends to collapse…
ZoomR: Memory Efficient Reasoning through Multi-Granularity Key Value Retrieval
David H. Yang, Yuxuan Zhu, Mohammad Mohammadi Amiri +4
Large language models (LLMs) have shown great performance on complex reasoning tasks but often require generating long intermediate thoughts before reaching a final answer. During…
The Alignment Game: A Theory of Long-Horizon Alignment Through Recursive Curation
Ali Falahati, Mohammad Mohammadi Amiri, Kate Larson +1
In self-consuming generative models that train on their own outputs, alignment with user preferences becomes a recursive rather than one-time process. We provide the first formal f…
Scalable, Explainable and Provably Robust Anomaly Detection with One-Step Flow Matching
Zhong Li, Qi Huang, Yuxuan Zhu +4
We introduce Time-Conditioned Contraction Matching (TCCM), a novel method for semi-supervised anomaly detection in tabular data. TCCM is inspired by flow matching, a recent generat…
Medix: Out-of-Distribution Detection from Unlabeled Wild Data via Robust Gradient Statistics
Momin Abbas, Ali Falahati, Hossein Goli +1
Out-of-distribution (OOD) detection plays a crucial role in ensuring the robustness of machine learning systems deployed in real-world applications. Recent approaches have explored…
OFMU: Optimization-Driven Framework for Machine Unlearning
Sadia Asif, Mohammad Mohammadi Amiri
Large language models deployed in sensitive applications increasingly require the ability to unlearn specific knowledge, such as user requests, copyrighted materials, or outdated i…
Toward Efficient Influence Function: Dropout as a Compression Tool
Yuchen Zhang, Mohammad Mohammadi Amiri
Assessing the impact the training data on machine learning models is crucial for understanding the behavior of the model, enhancing the transparency, and selecting training data. I…
Sparse Gradient Compression for Fine-Tuning Large Language Models
David H. Yang, Mohammad Mohammadi Amiri, Tejaswini Pedapati +2
Fine-tuning large language models (LLMs) for downstream tasks has become increasingly crucial due to their widespread use and the growing availability of open-source models. Howeve…
Disentangled Structural and Featural Representation for Task-Agnostic Graph Valuation
Ali Falahati, Mohammad Mohammadi Amiri
With the emergence of data marketplaces, the demand for methods to assess the value of data has increased significantly. While numerous techniques have been proposed for this purpo…