Back to Study Pods
AI & MLโ€ขPrimary Skill: PyTorch

AI & LLM Systems Pod

Collaborative cohort following Andrej Karpathy and Fast.ai roadmaps. We build fine-tuned small LLMs and RAG pipelines.

Pod Meeting Cadence
Tuesdays & Saturdays at 7:30 PM IST
Current Roadmap Milestone
Phase 3: Transformers & Attention Mechanism Deep Dive
Open Guide โ†’
Pod Readiness Score:0% Completed

Take turns: One student acts as interviewer while the other answers using the STAR technique (Situation, Task, Action, Result).

Q1Deep Learning Fundamentals

Explain the architectural difference between self-attention and cross-attention in Transformer models.

STAR Answer Strategy Tip:

Discuss queries, keys, and values projection matrices. Detail where encoder-decoder models inject cross-attention.

Q2Practical LLM Fine-Tuning

How do you mitigate catastrophic forgetting when fine-tuning an open-source model using LoRA / QLoRA?

STAR Answer Strategy Tip:

Explain rank matrices (A & B), parameter efficiency, freezing base weights, and validation loss tracking.

Q3Debugging & Problem Solving

Describe a project where you encountered vanishing or exploding gradients and how you diagnosed it.

STAR Answer Strategy Tip:

Use the STAR framework: Situation (training loss stagnated), Task (debug tensor values), Action (gradient clipping, layer norm), Result.

Q4RAG Evaluation (Ragas / TruLens)

How do you evaluate retrieval precision and context hallucination in a production RAG system?

STAR Answer Strategy Tip:

Mention context recall, context precision, answer relevancy, and embedding vector similarity thresholds.

Q5Industry Awareness & Motivation

What excites you most about open-weights AI models versus proprietary closed APIs?

STAR Answer Strategy Tip:

Highlight latency control, local data privacy, cost predictability, and independence from cloud vendor lock-in.