Anish DubeyinTowards Data ScienceTowards Monosemanticity: A Step Towards Understanding Large Language ModelsUnderstanding the mechanistic interpretability research problem and reverse-engineering these large language modelsJul 11Jul 11
Anish DubeyinTowards Data ScienceLLM Alignment: Reward-Based vs Reward-Free MethodsOptimization methods for LLM alignmentJul 51Jul 51
Anish DubeyinTowards AICompute-efficient Way to Scale LLM — Journey Around Data, Model, and ComputeScaling in AI models requires balancing model parameters, compute resources and dataJun 26Jun 26
Anish DubeyinTowards AIExploration of Parameters-efficient fine tuning methods (LoRA/MoRA/DoRA) in LLMParameter-efficient fine-tuning (PEFT) methods like LoRA, DoRA, and MoRA allow effective learning with fewer parametersJun 14Jun 14
Anish DubeyinTowards Data ScienceFlash attention(Fast and Memory-Efficient Exact Attention with IO-Awareness): A deep diveFlash attention is power optimization transformer attention mechanism that provides 15% efficiencyMay 29May 29
Anish DubeyStrategic analysis on the rise of spontaneous communication platformsSpontaneous communication is radicalizing the way we communicate or hangout remotely. Due to covid-19, there’s been a surge of people who…Jun 7, 2020Jun 7, 2020
Anish DubeyAmazon Echo vs Google HomeGoogle just unveiled its voice activated speaker powered by the Google assistant. Its a straight competition to Amazon echo which is…Oct 17, 2016Oct 17, 2016
Anish DubeyIs Netflix nearing the end of its glory days?Netflix has been one of our favorite online streaming services among Amazon Prime, Hulu and Netflix. Different streaming services focus on…Oct 9, 20161Oct 9, 20161