🔬 Applied Scientist & ML Specialist938 Problems · 127 Topics🌐 Linked to Tech Vault
TalentMe Science Depth: Universal Applied Science & ML Depth Problem Bank
938 end-to-end science depth, system design, transition bridge & BQ problems covering derivations, production trade-offs, pitfalls, and follow-ups.
🎯
Science Depth Mastery Tracker
0% MasteredLocally synced in your browser. Track, review, and master all 938 scientific & design topics.
Mastered: 0
Reviewing: 0
Unstarted: 938
🌐
Linked to TalentMe Tech Vault & MindmapsGlobal System
All 938 questions are directly linked to TalentMe's Tech Vault guides, atomic cards, and interactive mindmaps.
Found 24 in-depth questions
M3-057Deep Learning FoundationsGradient Vanishing & Explosion
MediumHow to Detect and Isolate Gradient Anomalies: Diagnostic Signals and Workflows
⚡ One-Sentence Answer
M3-058Deep Learning FoundationsGradient Vanishing & Explosion
HardGradient Conflict in Multi-Task Learning and Remediation Techniques
⚡ One-Sentence Answer
M3-059Deep Learning FoundationsGradient Vanishing & Explosion
MediumHow Residual Connections and Normalization Cooperate to Guarantee Gradient Flow
⚡ One-Sentence Answer
M3-060Deep Learning FoundationsGradient Vanishing & Explosion
HardThe Loss Spike Phenomenon: Gradient Norm Correlations, Root Causes, and Mitigations
⚡ One-Sentence Answer
M3-061Deep Learning FoundationsGradient Vanishing & Explosion
HardGradient Scale Mismatch in Embedding and Output Layers: Logit Scaling and Tying
⚡ One-Sentence Answer
M3-062Deep Learning FoundationsGradient Vanishing & Explosion
HardPer-Layer and Adaptive Gradient Clipping (AGC) Explained
⚡ One-Sentence Answer
M3-063Deep Learning FoundationsGradient Vanishing & Explosion
HardWhy Stochastic Gradient Noise Acts as Implicit Regularization in SGD
⚡ One-Sentence Answer
M3-064Deep Learning FoundationsLoss Functions & Objectives
EasyComparing MSE, MAE, and Huber Loss: Robustness and Target Optimality
⚡ One-Sentence Answer
M3-065Deep Learning FoundationsLoss Functions & Objectives
EasyCross-Entropy vs Binary Cross-Entropy (BCE) and Numerically Stable Implementations
⚡ One-Sentence Answer
M3-066Deep Learning FoundationsLoss Functions & Objectives
MediumFocal Loss: Design Motivation, Mathematical Derivation, and Modulating Factor
⚡ One-Sentence Answer
M3-067Deep Learning FoundationsLoss Functions & Objectives
MediumInfoNCE vs Triplet Loss in Contrastive Learning: Formulations and Information Capacity
⚡ One-Sentence Answer
M3-068Deep Learning FoundationsLoss Functions & Objectives
HardLoss Function and Evaluation Metric Mismatch: Causes, Examples, and Surrogates
⚡ One-Sentence Answer
M3-069Deep Learning FoundationsLoss Functions & Objectives
MediumKL Divergence in Distillation and Regularization: The Asymmetry Problem
⚡ One-Sentence Answer
M3-070Deep Learning FoundationsLoss Functions & Objectives
HardRanking Losses in Search and Recommendation: Pointwise, Pairwise, and Listwise
⚡ One-Sentence Answer
M3-071Deep Learning FoundationsTraining Stability & Mixed Precision
EasyFloating-Point Precisions in Deep Learning: FP16, BF16, and FP32 Trade-offs
⚡ One-Sentence Answer
M3-072Deep Learning FoundationsTraining Stability & Mixed Precision
MediumLoss Scaling Mechanism and Dynamic Adjustment in Mixed Precision
⚡ One-Sentence Answer
M3-073Deep Learning FoundationsTraining Stability & Mixed Precision
MediumSystematic Troubleshooting Workflow When Training Loss Becomes NaN
⚡ One-Sentence Answer
M3-074Deep Learning FoundationsTraining Stability & Mixed Precision
HardOperators Most Prone to Instability Under Low-Precision (FP16/BF16)
⚡ One-Sentence Answer
M3-075Deep Learning FoundationsTraining Stability & Mixed Precision
HardGradient Checkpointing (Activation Rematerialization): Memory vs Compute Trade-offs
⚡ One-Sentence Answer
M3-076Deep Learning FoundationsTraining Stability & Mixed Precision
MediumAttention Numerical Stability and How FlashAttention Solves It
⚡ One-Sentence Answer
M3-077Deep Learning FoundationsTraining Stability & Mixed Precision
MediumWhy LayerNorm and Softmax Must Retain FP32 Precision in Mixed-Precision Training
⚡ One-Sentence Answer
M3-078Deep Learning FoundationsDistributed Training Basics
EasyComparing Distributed Parallelism: Data (DP), Tensor (TP), and Pipeline (PP)
⚡ One-Sentence Answer
M3-079Deep Learning FoundationsDistributed Training Basics
EasyZeRO (Zero Redundancy Optimizer): The Three Sharding Stages Explained
⚡ One-Sentence Answer
M3-080Deep Learning FoundationsDistributed Training Basics
MediumCollective Communication Primitives: All-Reduce, All-Gather, and Reduce-Scatter
⚡ One-Sentence Answer