ML @ CMU

SEPTEMBER 15, 2023

TLDR: Current SOTA methods for scene understanding, though impressive, often fail to decompose out-of-distribution scenes. In our ICML paper, Slot-TTA ( [link] ) we find that optimizing per test sample over reconstruction loss improves scene decomposition accuracy. Problem Statement: In machine learning, we often assume the train and test split are IID samples from the same distribution.

Machine Learning 211

Machine Learning Machine Learning Supervised Learning Algorithm 211

ML @ CMU

JUNE 30, 2023

Empirical study: We evaluated three approaches for robots to navigate to objects in six visually diverse homes. TLDR : Semantic navigation is necessary to deploy mobile robots in uncontrolled environments like our homes, schools, and hospitals. Many learning-based approaches have been proposed in response to the lack of semantic understanding of the classical pipeline for spatial navigation.

Analytics 222

Analytics Analytics Machine Learning Machine Learning 222

ML @ CMU

JUNE 5, 2023

ReLM enables writing tests that are guaranteed to come from the set of valid strings, such as dates. Without ReLM, LLMs are free to complete prompts with non-date answers, which are difficult to assess. TL;DR: While large language models (LLMs) have been touted for their ability to generate natural-sounding text, there are concerns around potential negative effects of LLMs such as data memorization, bias, and inappropriate language.

Machine Learning 154

Machine Learning Machine Learning 154

ML @ CMU

MAY 12, 2023

TL;DR: We study the use of differential privacy in personalized, cross-silo federated learning ( NeurIPS’22 ), explain how these insights led us to develop a 1st place solution in the US/UK Privacy-Enhancing Technologies (PETs) Prize Challenge , and share challenges and lessons learned along the way. If you are feeling adventurous, checkout the extended version of this post with more technical details!

Data Silos 130

Data Silos Algorithm ML ML 130

ML @ CMU

APRIL 14, 2023

Example of embodied commonsense reasoning. A robot proactively identifies a remote on the floor and knows it is out of place without instruction. Then, the robot figures out where to place it in the scene and manipulates it there. For robots to operate effectively in the world, they should be more than explicit step-by-step instruction followers. Robots should take actions in situations when there is a clear violation of the normal circumstances and be able to infer relevant context from partial

Artificial Intelligence 363

Artificial Intelligence Artificial Intelligence Machine Learning Machine Learning 363

ML @ CMU

MARCH 31, 2023

Figure 1. This blog post discusses the effectiveness of black-box model explanations in aiding end users to make decisions. We observe that explanations do not in fact help with concrete applications such as fraud detection and paper matching for peer review. Our work further motivates novel directions for developing and evaluating tools to support human-ML interactions.

ML 246

ML ML Algorithm Machine Learning 246

ML @ CMU

MARCH 30, 2023

Figure 1. This blog post discusses the effectiveness of black-box model explanations in aiding end users to make decisions. We observe that explanations do not in fact help with concrete applications such as fraud detection and paper matching for peer review. Our work further motivates novel directions for developing and evaluating tools to support human-ML interactions.

ML 130

ML ML Algorithm Machine Learning 130

ML @ CMU

MARCH 24, 2023

Figure 1: Behavior-driven AI development centers model iteration on evaluating and improving specific real-world use cases. It has never been easier to prototype AI-driven systems. With a bit of programming knowledge and a couple of hours, you can spin up a chatbot for your notes , a text-based image editor , or a tool for summarizing customer feedback.

AI 362

AI AI Python ML 362

ML @ CMU

FEBRUARY 24, 2023

Figure 1 : Overview of RL Prompt for discrete prompt optimization. All language models (LMs) are frozen. We build our policy network by training a task-specific multi-layer perceptron (MLP) network inserted into a frozen pre-trained LM. The figure above illustrates 1) generation of a prompt ( left ), 2) example usages in a masked LM for classification ( top right ) and a left-to-right LM for generation ( bottom right ), and 3) update of the MLP using RL reward signals ( red arrows ).

Natural Language Processing 253

Natural Language Processing Algorithm Computer Science Computer Science 253

ML @ CMU

JANUARY 13, 2023

We perform open vocabulary detection of the objects mentioned in the sentence using both bottom-up and top-down feedback. Object detection is the fundamental computer vision task of finding all “objects” that are present in a visual scene. However, this raises the question, what is an object? Typically, this question is side-stepped by defining a vocabulary of categories and then training a model to detect instances of this vocabulary.

Machine Learning 130

Machine Learning Machine Learning 130