Latest posts tagged with #Debiasing on Bluesky
Say My Name: a Model's Bias Discovery Framework
Massimiliano Ciranni, Luca Molinaro, Carlo Alberto Barbano et al.
Action editor: Sungsoo Ahn
https://openreview.net/forum?id=EuUxiiDZ7d
#bias #biases #debiasing
Controlled Model Debiasing through Minimal and Interpretable Updates
Federico Di Gennaro, Thibault Laugel, Vincent Grari, Marcin Detyniecki
Action editor: Shahin Jabbari
https://openreview.net/forum?id=B9fdU4qjpD
#fairness #adversarial #debiasing
Fairness and Disentanglement: A Critical Review of Predominant Bias in Neural Networks
Jiazhi Li, Mahyar Khayatkhoei, Jiageng Zhu et al.
Action editor: Stanislaw Jastrzebski
https://openreview.net/forum?id=LLiJ1WsL2e
#bias #biases #debiasing
Debiasing Heuristics Improve Low-Resource Web-Mined Parallel Corpora
Researchers added model‑agnostic debiasing heuristics to clean web‑mined parallel sentences, improving low‑resource machine‑translation performance. Code and data are on arXiv. getnews.me/debiasing-heuristics-imp... #debiasing #machinetranslation
Universal Debiasing Framework Boosts Fairness in LLM Tabular Data
Debiasing framework for LLM tabular data offers UDF‑DPO and UDF‑MIX, reducing info between advantaged and protected groups and delivering fairness trade‑offs over baseline generators. Read more: getnews.me/universal-debiasing-fram... #llm #debiasing
Debiased Learning via Neural Collapse: Avoiding Shortcut Models
A new 'avoid-shortcut' framework aligns training with Neural Collapse without extra steps, and the paper's second preprint version was posted in September 2025. getnews.me/debiased-learning-via-ne... #neuralcollapse #debiasing
New Low‑Variance De‑biasing Method Boosts Learning from Label Proportions
New low‑variance de‑biasing method for Learning from Label Proportions works with losses and delivers sample‑complexity bounds, cutting training samples. Read more: getnews.me/new-low-variance-de-bias... #learningfromlabelproportions #debiasing
A 2021 paper in a pay-to-publish #Nature journal claimed #mindfulness tasks (notice 3 things around you or differences between 2 similar images) reduced some cognitive #biases: doi.org/10.1057/s415...
The mindful #debiasing was undetected in a re-analysis and 2 replications: doi.org/10.1007/s126...
Methods
Results
Does awareness of intuition's fallibility help people avoid faulty intuitions?
#Teaching students #DualProcessTheory didn't help them avoid faulty intuitions about #physics problems.
doi.org/10.1103/Phys...
#edu #cogSci #bias #debiasing #psychology #epistemology #rationality
#Creativity Has Left the Chat: The Price of #Debiasing Language Models
arxiv.org/abs/2406.05587
#AI #LLM #OpenAI #ChatGPT #Gemini
