Skip to main content
← Choose a different target

Unlock: GPT Series Evolution

The progression from GPT-1 (117M) through GPT-4, the o-series, GPT-5, GPT-5.4, gpt-oss, and GPT-Rosalind: what changed through scale, post-training, reasoning-time compute, tool use, and domain specialization.

391 Prerequisites0 Mastered0 Working264 Gaps
Prerequisite mastery32%
Recommended probe

Universal Approximation Theorem is your weakest prerequisite with available questions. You haven't been assessed on this topic yet.

GPT Series EvolutionTARGET
Universal Approximation TheoremCoreWEAKEST
Not assessed5 questions
Forgetting Transformer (FoX)Research
No quiz
Sparse Recovery and Compressed SensingResearch
No quiz
Scaling LawsResearch
Not assessed3 questions
Attention Mechanism TheoryResearch
Not assessed11 questions
BERT and the Pretrain-Finetune ParadigmResearch
Not assessed3 questions
Post-Training OverviewFrontier
No quiz
Prompt Engineering and In-Context LearningFrontier
No quiz
RLHF and AlignmentResearch
Not assessed3 questions
Transformer ArchitectureResearch
Not assessed11 questions
Tokenization and Information TheoryResearch
Not assessed6 questions

Sign in to track your mastery and see personalized gap analysis.