159 citations · 178 across the 17 of their papers we have counts for
17 papers
Stage-Level Executor Allocation in Apache Spark with Cost-Performance Trade-offs
Miriam Rateike, Isaac Waweru Wambugu, Celia Cintas +3
Allocating executors (i.e. compute resources) to distributed processing systems must balance resource costs of scaling-out unnecessarily against artificial, performance-limiting bo…
Retrieval with Multiple Query Vectors through Anomalous Pattern Detection
Allassan Tchangmena A Nken, Baimam Boukar Jean Jacques, Miriam Rateike +2
A classical vector retrieval problem typically considers a \emph{single} query embedding vector as input and retrieves the most similar embedding vectors from a vector database. Ho…
Calibrating Lightweight Sparse Autoencoder Feature Steering
Ananya Joshi, Celia Cintas, Skyler Speakman
Sparse autoencoders (SAEs) can enable inference-time topic steering by modifying latent feature activations, but existing steering methods often fail when target-aligned features a…
Localizing Persona Representations in LLMs
Celia Cintas, Miriam Rateike, Erik Miehling +2
We present a study on how and where personas -- defined by distinct sets of human characteristics, values, and beliefs -- are encoded in the representation space of large language…
Efficient Representation of the Activation Space in Deep Neural Networks
Tanya Akumu, Celia Cintas, Girmaw Abebe Tadesse +3
The representations of the activation space of deep neural networks (DNNs) are widely utilized for tasks like natural language processing, anomaly detection and speech recognition.…
Weakly Supervised Detection of Hallucinations in LLM Activations
Miriam Rateike, Celia Cintas, John Wamburu +2
We propose an auditing method to identify whether a large language model (LLM) encodes patterns such as hallucinations in its internal states, which may propagate to downstream tas…