🔬 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: 136
🌐
Linked to TalentMe Tech Vault & MindmapsM5: NLP & Large Language Models
All 938 questions are directly linked to TalentMe's Tech Vault guides, atomic cards, and interactive mindmaps.
Found 24 in-depth questions
M5-025NLP & Large Language ModelsInstruction Tuning & Supervised Fine-Tuning
HardDivision of Labor Between SFT and RLHF/DPO
⚡ One-Sentence Answer
M5-026NLP & Large Language ModelsInstruction Tuning & Supervised Fine-Tuning
MediumChoosing Learning Rates and Epochs in SFT
⚡ One-Sentence Answer
M5-027NLP & Large Language ModelsInstruction Tuning & Supervised Fine-Tuning
MediumChat Template Formatting and Multi-Turn Conversation Construction
⚡ One-Sentence Answer
M5-028NLP & Large Language ModelsInstruction Tuning & Supervised Fine-Tuning
HardRejection Sampling Fine-Tuning (RFT / STaR) Mechanisms
⚡ One-Sentence Answer
M5-029NLP & Large Language ModelsInstruction Tuning & Supervised Fine-Tuning
MediumSFT Data Sources and Engineering Trade-offs
⚡ One-Sentence Answer
M5-030NLP & Large Language ModelsAlignment & RLHF
EasyThe Three-Stage RLHF Pipeline
⚡ One-Sentence Answer
M5-031NLP & Large Language ModelsAlignment & RLHF
EasyBradley-Terry Preference Model and Reward Model Loss Formulation
⚡ One-Sentence Answer
M5-032NLP & Large Language ModelsAlignment & RLHF
MediumPPO Clipped Surrogate Objective and the Rationale for Min and Clip
⚡ One-Sentence Answer
M5-033NLP & Large Language ModelsAlignment & RLHF
MediumGeneralized Advantage Estimation (GAE) and Advantage in RLHF
⚡ One-Sentence Answer
M5-034NLP & Large Language ModelsAlignment & RLHF
HardReward Hacking and Mitigation Strategies in RLHF
⚡ One-Sentence Answer
M5-035NLP & Large Language ModelsAlignment & RLHF
HardEngineering Challenges and Stability Issues in RLHF
⚡ One-Sentence Answer
M5-036NLP & Large Language ModelsAlignment & RLHF
MediumThe Role and Dynamics of the KL Penalty Coefficient in RLHF
⚡ One-Sentence Answer
M5-037NLP & Large Language ModelsAlignment & RLHF
HardReward Model Generalization and Over-Optimization
⚡ One-Sentence Answer
M5-038NLP & Large Language ModelsAlignment & RLHF
HardOn-Policy Sampling vs. Off-Policy Data in RLHF
⚡ One-Sentence Answer
M5-039NLP & Large Language ModelsAlignment & RLHF
HardPPO Implementation Nuances in LLMs: Token-Level vs. Sequence-Level
⚡ One-Sentence Answer
M5-040NLP & Large Language ModelsAlignment & RLHF
HardMotivations and Practices of Iterative / Online RLHF
⚡ One-Sentence Answer
M5-041NLP & Large Language ModelsAlignment & RLHF
MediumConstitutional AI and RLAIF (Reinforcement Learning from AI Feedback)
⚡ One-Sentence Answer
M5-042NLP & Large Language ModelsDPO Family (DPO / KTO / ORPO)
EasyFull Mathematical Derivation of Direct Preference Optimization (DPO) Loss
⚡ One-Sentence Answer
M5-043NLP & Large Language ModelsDPO Family (DPO / KTO / ORPO)
EasySystematic Comparison of PPO, DPO, and GRPO
⚡ One-Sentence Answer
M5-044NLP & Large Language ModelsDPO Family (DPO / KTO / ORPO)
MediumMotivations and Distinctions of ORPO and KTO
⚡ One-Sentence Answer
M5-045NLP & Large Language ModelsDPO Family (DPO / KTO / ORPO)
MediumSimPO: Simple Preference Optimization Improvements
⚡ One-Sentence Answer
M5-046NLP & Large Language ModelsDPO Family (DPO / KTO / ORPO)
HardFailure Modes and Fundamental Limitations of DPO
⚡ One-Sentence Answer
M5-047NLP & Large Language ModelsDPO Family (DPO / KTO / ORPO)
HardPreference Data Construction and Quality Control
⚡ One-Sentence Answer
M5-048NLP & Large Language ModelsDPO Family (DPO / KTO / ORPO)
MediumIdentity Preference Optimization (IPO) Motivation and Regularization
⚡ One-Sentence Answer