Stop just using AI—learn to hack it! 🤖🛠️ Join "Machine Unlearning for Artists" to demystify the AI black box. We’ll explore hidden mechanics, expose biases, and learn to creatively manipulate machine learning models.
simonweckert.com/machineunlea...
#MachineUnlearning #AIEthics #AIBias #HackAI

We would like to introduce the new publication in TAI, application of Thirty-Six Stratagems (三十六计) in medical image analysis, by Prof. Dapeng Wu, from City University of Hong Kong

www.sciltp.com/journals/tai...

#federatedlearning #medicalimageanalysis #trustworthymachinelearning #machineunlearning

#AI #Privacy #MachineLearning #LargeLanguageModels #LLMs #Research #ArtificialIntelligence #DataSecurity #DifferentialPrivacy #MachineUnlearning #TMLR #Publication #HealthcareAI #DataTools4Heart #Translated

Today at #AISTATS2026

An Illusion of Unlearning? Assessing #MachineUnlearning Through Internal Representations.

MU algorithms seem to achieve perfect unlearning of classes through misalignment of weights & features rather than truly unlearning. We analyze this through #NeuralCollapse

See you at #EACL2026 in Rabat 🕌!

#LLMs #AI #NLP #NLProc #MachineUnlearning #AIPrivacy #AISafety #ExplainableAI #ResponsibleAI @cs-tudarmstadt.bsky.social

🚨New preprint alert! “Supporting Trustworthy AI Through Machine Unlearning” now available on SSRN: papers.ssrn.com/sol3/papers.... 
Curious about how new Machine Unlearning techniques can make AI better? Summary 🧵below…
#HCI #STS #AI #MachineUnlearning #MachineLearning

At NeurIPS 2025, researchers tested machine‑unlearning on a text‑to‑music model, showing it can forget specific training data while keeping overall quality. Read more: https://getnews.me/machine-unlearning-offers-opt-out-solution-for-ai-music-generation/ #machineunlearning #aimusic

A new study shows a constraint‑driven weight tweak that blocks targeted vocabularies and improves jailbreak resistance, beating max‑min methods without external classifier. Read more: https://getnews.me/machine-unlearning-robustness-via-constrained-llm-interventions/ #machineunlearning #robustness

Machine unlearning reduces the success of a one‑word safety‑mirage attack by up to 60.17% and cuts unnecessary rejections by over 84.20%. Read more: https://getnews.me/machine-unlearning-cuts-safety-mirage-in-vision-language-models/ #machineunlearning #aialignment

OFMU employs bi‑level optimization with a penalty to separate forgetting and retention, delivering convergence for large language models. Sept 2025. Read more: https://getnews.me/ofmu-framework-boosts-machine-unlearning-for-ai-models/ #machineunlearning #ai #gdpr

TraceHiding offers a scalable, importance‑aware way to delete mobility trajectories from models without retraining, achieving up to 40× speed‑up, tested on HO‑Rome, HO‑Geolife and HO‑NYC. https://getnews.me/tracehiding-scalable-machine-unlearning-for-mobility-data/ #machineunlearning #privacy

CoUn uses contrastive learning on retained data to push forgotten samples away in latent space, boosting machine‑unlearning effectiveness without full retraining. Read more: https://getnews.me/contrastive-learning-boosts-machine-unlearning-effectiveness/ #machineunlearning #contrastivelearning #ai

Peut-on imposer une « amnésie sélective » à l’ #IAgénérative et pourquoi ?
#désapprentissage #MachineUnlearning
www.linkedin.com/posts/fran%C...

When people from Machine learning say that something is too "statistical" I want to vomit #machineUNlearning

🎧 Listen now: youtu.be/osHrYXwCvpQ
📢 Follow us for more stories from women shaping the future of AI.
#WiAIR #WomenInAI #LLMs #ResponsibleAI #AIinHealthcare #Interpretability #BiasInAI #MachineUnlearning #AIethics