What are alternatives to CLIP (SigLIP, EVA-CLIP)?
Machine Learning / AI
What is DoRA and how does it differ from LoRA?
What is CBAM (Convolutional Block Attention Module)?
How does BERT capture the semantic similarity between sentences that are written differently but have the same meaning?
What is diverse beam search?
What do the TF and IDF components mean?
How does top-k differ from top-p (nucleus) sampling?
How to reduce the size of KV-cache (GQA, MQA, sliding window)?
When does deep learning outperform GBDT on tabular data?
Why is RAG better than fine-tuning for a QA task on internal documentation?
What are the features of job and resume ranking?
Tell us about ALBERT and its parameter sharing between layers.
What is negative sampling in training word2vec and why is it needed?
What are consistency models (LCM, SDXL-Lightning)?
What is prompt injection and what attacks have you seen in practice?
What is do-calculus Pearl?
What is InstructBLIP?
What is target network in DQN?
How accurate is data anonymization?
What is pose estimation (top-down vs bottom-up)? What is OpenPose?