Novel Concept

Research on novel concepts spans diverse fields, focusing on developing methods for their efficient discovery, representation, and utilization. Current efforts involve leveraging advanced machine learning models, including large language models and pre-trained transformers, to generate and integrate new concepts into existing knowledge bases or systems, as well as adapting algorithms like those used in reinforcement learning and distributionally robust optimization to improve efficiency and robustness. This work has significant implications for various domains, from improving ontology construction and robotic exploration to enhancing the capabilities of text-to-image models and advancing the development of robust machine learning systems.

Papers

November 1, 2024

Decoding Dark Matter: Specialized Sparse Autoencoders for Interpreting Rare Concepts in Foundation Models
Aashiq Muhamed, Mona Diab, Virginia Smith
Foundation Model Supervised Autoencoder Many Sparse Sparsity Increase Dense Retrieval Machine Translated Sparse Autoencoders Dark Matter Novel Concept

October 29, 2024

Rare-to-Frequent: Unlocking Compositional Generation Power of Diffusion Models on Rare Concepts with LLM Guidance
Dongmin Park, Sebin Kim, Taehong Moon, Minkyu Kim, Kangwook Lee, Jaewoong Cho
Diffusion Explainer Semantic Affinity LLM Policy Novel Concept Diffusion Inference Compositional Generation Pre Trained Diffusion Multi Concept Composition

October 24, 2024

ConceptDrift: Uncovering Biases through the Lens of Foundational Models
Cristian Daniel Păduraru, Antonio Bărbălau, Radu Filipescu, Andrei Liviu Nicolicioiu, Elena Burceanu
Pre Trained Model Camera Lens Topic Bias Text Datasets Foundational Model Novel Concept

October 17, 2024

Chain of Ideas: Revolutionizing Research Via Novel Idea Development with LLM Agents
Long Li, Weiwen Xu, Jiayan Guo, Ruochen Zhao, Xingxuan Li, Yuqian Yuan, Boqiang Zhang, Yuming Jiang, Yifei Xin, Ronghao Dang, Deli Zhao, Yu Rong, Tian Feng, Lidong Bing
DH Research LLM Agent Side Chain Human Generated Idea Idea Generation Novel Concept Research Task

October 1, 2024

Beyond Minimax Rates in Group Distributionally Robust Optimization via a Novel Notion of Sparsity
Quan Nguyen, Nishant A. Mehta, Cristóbal Guzmán
Sample Complexity Sparsity Increase Robust Optimization Distributionally Robust Optimization Novel Concept Minimax Rate

February 27, 2024

A Language Model based Framework for New Concept Placement in Ontologies
Hang Dong, Jiaoyan Chen, Yuan He, Yongsheng Gao, Ian Horrocks
Language Model New Framework Pre Trained Language Model Top Level Ontology Multi Label Image Novel Concept Snomed Ct

November 15, 2023

A novel concept for Titan robotic exploration based on soft morphing aerial robots
Fernando Ruiz, Begona Arrue, Anibal Ollero
Aerial Robot Autonomous Exploration Elastomer Deformation Novel Concept

October 12, 2023

SingleInsert: Inserting New Concepts from a Single Image into Text-to-Image Models for Flexible Editing
Zijie Wu, Chaohui Yu, Zhen Zhu, Fan Wang, Xiang Bai
Image Generation Text to Image Model Text to Image Single Image Image to Text Object Insertion Controllable Editing Novel Concept Concept Generation

August 10, 2023

Normalized Gradients for All
Francesco Orabona
Black Box Natural Gradient Short Note Novel Concept Gradient Normalization

June 26, 2023

Ontology Enrichment from Texts: A Biomedical Dataset for Concept Discovery and Placement
Hang Dong, Jiaoyan Chen, Yuan He, Ian Horrocks
Top Level Ontology Knowledge Base Text Based Concept Discovery Biomedical Datasets Global Placement Novel Concept Semantic Enrichment

May 23, 2023

Negative Feedback Training: A Novel Concept to Improve Robustness of NVCIM DNN Accelerators
Yifan Qin, Zheyu Yan, Wujie Wen, Xiaobo Sharon Hu, Yiyu Shi
Native Robustness DNN Accelerator DNN Inference Non Volatile Memory DNN Robustness Novel Concept Device Variation Negative Training

April 27, 2023

Generating images of rare concepts using pre-trained diffusion models
Dvir Samuel, Rami Ben-Ari, Simon Raviv, Nir Darshan, Gal Chechik
Image Generation Text to Image Diffusion Model Text to Image Model Pre Trained Diffusion Model Underrepresented Group Novel Concept Shot Classification Benchmark Imbalanced Training

April 1, 2023

From Conception to Deployment: Intelligent Stroke Prediction Framework using Machine Learning and Performance Evaluation
Leila Ismail, Huned Materwala
Machine Learning Random Forest Performance Evaluation Software Deployment Stroke Detection Stroke Prediction Novel Concept Stroke Data

March 28, 2023

Large-scale Pre-trained Models are Surprisingly Strong in Incremental Novel Class Discovery
Mingxuan Liu, Subhankar Roy, Zhun Zhong, Nicu Sebe, Elisa Ricci
LeArning Abstract Full Model Large Scale Lifelong Learning Large Scale Pre Trained Model Novel Class Discovery Novel Concept

December 28, 2022

Towards Learning Abstractions via Reinforcement Learning
Erik Jergéus, Leo Karlsson Oinonen, Emil Carlsson, Moa Johansson
Reinforcement Learning Multi Agent System Symbolic Approach Novel Concept Abstraction Learning Neural Symbolic

September 30, 2022

Rethinking Data Heterogeneity in Federated Learning: Introducing a New Notion and Standard Benchmarks
Mahdi Morafah, Saeed Vahidian, Chen Chen, Mubarak Shah, Bill Lin
Pseudo Label Data Heterogeneity Model Aggregation Statistical Heterogeneity Training Data Distribution Novel Concept Human Aligned Benchmark

August 24, 2022

ForestEyes Project: Conception, Enhancements, and Challenges
Fernanda B. J. R. Dallaqua, Álvaro Luiz Fazenda, Fabio A. Faria
Technical Challenge Feature Enhancement Citizen Science Novel Concept Deforestation Area Deforestation Detection Forest Mapping

April 2, 2022

Production of Categorical Data Verifying Differential Privacy: Conception and Applications to Machine Learning
Héber H. Arcolezi
Machine Learning Financial Application Differential Privacy Privacy Preserving Production Incident Private Machine Learning Privacy Utility Trade Novel Concept

November 16, 2021

Generative Pre-Trained Transformer for Design Concept Generation: An Exploration
Qihao Zhu, Jianxi Luo
Environment Exploration Generative Pre Trained Transformer Generative Design Novel Concept Design Concept Generation

Novel Concept

Papers

Decoding Dark Matter: Specialized Sparse Autoencoders for Interpreting Rare Concepts in Foundation Models

Rare-to-Frequent: Unlocking Compositional Generation Power of Diffusion Models on Rare Concepts with LLM Guidance

ConceptDrift: Uncovering Biases through the Lens of Foundational Models

Chain of Ideas: Revolutionizing Research Via Novel Idea Development with LLM Agents

Beyond Minimax Rates in Group Distributionally Robust Optimization via a Novel Notion of Sparsity

A Language Model based Framework for New Concept Placement in Ontologies

A novel concept for Titan robotic exploration based on soft morphing aerial robots

SingleInsert: Inserting New Concepts from a Single Image into Text-to-Image Models for Flexible Editing

Normalized Gradients for All

Ontology Enrichment from Texts: A Biomedical Dataset for Concept Discovery and Placement

Negative Feedback Training: A Novel Concept to Improve Robustness of NVCIM DNN Accelerators

Generating images of rare concepts using pre-trained diffusion models

From Conception to Deployment: Intelligent Stroke Prediction Framework using Machine Learning and Performance Evaluation

Large-scale Pre-trained Models are Surprisingly Strong in Incremental Novel Class Discovery

Towards Learning Abstractions via Reinforcement Learning

Rethinking Data Heterogeneity in Federated Learning: Introducing a New Notion and Standard Benchmarks

ForestEyes Project: Conception, Enhancements, and Challenges

Production of Categorical Data Verifying Differential Privacy: Conception and Applications to Machine Learning

Generative Pre-Trained Transformer for Design Concept Generation: An Exploration