![[2606.02814] Do Neural Retrievers Prefer Certain Documents? Evidence of Learned Relevance Priors](/static/img/curated/desks/afbytes-tech.png)
[2606.02814] Do Neural Retrievers Prefer Certain Documents? Evidence of Learned Relevance Priors
Abstract page for arXiv paper 2606.02814: Do Neural Retrievers Prefer Certain Documents? Evidence of Learned Relevance Priors
America Forever Bytes
bias · AFBytes
Other
bias
Recent coverage
[2606.01914] Mechanistic Diagnostics of Spatial Lexical Bias in Multimodal Large Language Model Spatial Reasoning
Abstract page for arXiv paper 2606.01914: Mechanistic Diagnostics of Spatial Lexical Bias in Multimodal Large Language Model Spatial Reasoning
[2606.01926] Mitigating Bias in Locally Constrained Decoding via Tractable Proposals
Abstract page for arXiv paper 2606.01926: Mitigating Bias in Locally Constrained Decoding via Tractable Proposals
[2606.02339] Entropy Minimization without Model Collapse: Mitigating Prediction Bias in Medical Imaging
Abstract page for arXiv paper 2606.02339: Entropy Minimization without Model Collapse: Mitigating Prediction Bias in Medical Imaging
Feedback Bias? How AI Adjusts Replies Based on Race and Gender, Research Finds
Feedback Bias? How AI Adjusts Replies Based on Race and Gender, Research Finds https://www.kqed.org/mindshift/66299/feedback-bias-how-ai-adjusts-replies-based-o...
Don’t throw the ‘lawfare’ dart at an attempt to counter school bias
Legal action has not been the first resort for most Jewish parents. It comes after repeated complaints, meetings and requests for intervention fail to produce a...
Can Digital Tools Fix Bias in Mental Health Triage?
Recent advances in digital technologies hold both promise and potential pitfalls for mental health triage. In this News and Perspectives article, JMIR Correspon...
Jumping to conclusions
Carelessly jumping to a conclusion is a typical consequence of failing to properly analyze new information.
[2605.28098] Examining Agents' Bias Amplification versus Suppression in Multi-Agent Systems
Abstract page for arXiv paper 2605.28098: Examining Agents' Bias Amplification versus Suppression in Multi-Agent Systems
‘SNL’ can’t stop making fun of conservatives, Trump — and unbelievable data proves it
A whopping 91% of jokes delivered on “Saturday Night Live’s” Weekend Update segment this season targeted conservatives, while 82% of characters during the...