649 citations · 689 across the 9 of their papers we have counts for
9 papers
How Language Models Organize and Structure Moral Knowledge
Orion Reblitz-Richardson
How do large language models (LLMs) organize moral knowledge? Models detect moral content broadly, but detection is a low bar. We ask whether they go further, distinguishing moral…
Output Dilution: Redundant but Fragile Representations in MoE Models
Orion Reblitz-Richardson
Mixture-of-Experts (MoE) models appear to encode moral content as robustly as dense models, yet prove far more fragile in their encoding. In OLMoE-1B-7B, linear probes recover mora…
When Probing Accuracy Saturates, Fragility Resolves: A Complementary Metric for LLM Pre-Training Analysis
Orion Reblitz-Richardson
Standard linear probing declares a property "encoded" when a classifier on hidden states achieves high accuracy. The protocol works well on a snapshot but breaks across pre-trainin…
ExecuTorch -- A Unified PyTorch Solution to Run AI Models On-Device
Mergen Nachin, Digant Desai, Sicheng Stephen Jia +36
Local execution of AI on edge devices is important for low latency and offline operation. However, deploying models on diverse hardware remains fragmented, often requiring model co…
A Tour of Visualization Techniques for Computer Vision Datasets
Bilal Alsallakh, Pamela Bhattacharya, Vanessa Feng +4
We survey a number of data visualization techniques for analyzing Computer Vision (CV) datasets. These techniques help us understand properties and latent patterns in such data, by…
Investigating sanity checks for saliency maps with image and text classification
Narine Kokhlikyan, Vivek Miglani, Bilal Alsallakh +2
Saliency maps have shown to be both useful and misleading for explaining model predictions especially in the context of images. In this paper, we perform sanity checks for text mod…
Investigating Saturation Effects in Integrated Gradients
Vivek Miglani, Narine Kokhlikyan, Bilal Alsallakh +2
Integrated Gradients has become a popular method for post-hoc model interpretability. De-spite its popularity, the composition and relative impact of different regions of the integ…
Mind the Pad -- CNNs can Develop Blind Spots
Bilal Alsallakh, Narine Kokhlikyan, Vivek Miglani +2
We show how feature maps in convolutional networks are susceptible to spatial bias. Due to a combination of architectural choices, the activation at certain locations is systematic…
Captum: A unified and generic model interpretability library for PyTorch
Narine Kokhlikyan, Vivek Miglani, Miguel Martin +8
In this paper we introduce a novel, unified, open-source model interpretability library for PyTorch [12]. The library contains generic implementations of a number of gradient and p…