Noisy Explanation

Noisy explanation research focuses on improving the reliability and interpretability of explanations generated by machine learning models, particularly in scenarios with imperfect data or complex reasoning processes. Current efforts concentrate on understanding and mitigating the impact of noise in explanations, exploring techniques like improved loss functions for training translation models, and developing algorithms that can handle noisy labels without requiring prior knowledge of noise distributions. This work is crucial for building trust in AI systems and ensuring the responsible use of machine learning in various applications, from natural language processing to computer vision.

Papers

July 1, 2024

Deciphering the Factors Influencing the Efficacy of Chain-of-Thought: Probability, Memorization, and Noisy Reasoning
Akshara Prabhakar, Thomas L. Griffiths, R. Thomas McCoy
Chain of Thought Limited Memorization Probability Pr(E1 Multi Step Reasoning Capability Symbolic Reasoning Task World War II Era Cipher Noisy Explanation

October 23, 2023

Reference Free Domain Adaptation for Translation of Noisy Questions with Question Specific Rewards
Baban Gain, Ramakrishna Appicharla, Soumya Chennabasavaraj, Nikesh Garera, Asif Ekbal, Muthusamy Chelliah
Neural Machine Translation View Translation Parallel Corpus Masked Language Modeling Reward Structure Noisy Explanation

April 13, 2023

Bayes classifier cannot be learned from noisy responses with unknown noise rates
Soham Bakshi, Subha Maity
Noisy Label Label Noise Simple Classifier Classification Problem Bayes Classifier Noisy Explanation

March 16, 2023

Empowering CAM-Based Methods with Capability to Generate Fine-Grained and High-Faithfulness Explanations
Changqing Qiu, Fusheng Jin, Yining Zhang
Fine Grained NCD Method Capability Evolution Explainability Problem Explanation Faithfulness Noisy Explanation

February 5, 2022

A Game-theoretic Understanding of Repeated Explanations in ML Models
Kavita Kumari, Murtuza Jadliwala, Sumit Kumar Jha, Anindya Maiti
Machine Learning Machine Learning Model Human Prediction Game Theoretic Game Theory Continuous Time Stochastic Noisy Explanation

November 19, 2021

A Worker-Task Specialization Model for Crowdsourcing: Efficient Inference and Fundamental Limits
Doyeon Kim, Jeonghwan Lee, Hye Won Chung
Efficient Inference Crowdsourcing Context Data Labeling Fundamental Limit Task Specialization Crowdsourcing System Noisy Explanation

Noisy Explanation

Papers

Deciphering the Factors Influencing the Efficacy of Chain-of-Thought: Probability, Memorization, and Noisy Reasoning

Reference Free Domain Adaptation for Translation of Noisy Questions with Question Specific Rewards

Bayes classifier cannot be learned from noisy responses with unknown noise rates

Empowering CAM-Based Methods with Capability to Generate Fine-Grained and High-Faithfulness Explanations

A Game-theoretic Understanding of Repeated Explanations in ML Models

A Worker-Task Specialization Model for Crowdsourcing: Efficient Inference and Fundamental Limits