Unlock: Hallucination Theory
Why language models confabulate, framed by the Kalai-Nachum-Vempala-Zhang (2025) statistical lower bound on pretraining error and the evaluation-incentive equilibrium that prevents post-training from removing it. Calibration, conformal prediction, structural failure modes (reversal curse, snowballing), and the measurement gap (FActScore, SAFE).
304 Prerequisites0 Mastered0 Working237 Gaps
Prerequisite mastery22%
Recommended probe
Floating-Point Arithmetic is your weakest prerequisite with available questions. You haven't been assessed on this topic yet.
Hallucination TheoryTARGET
Not assessed3 questions
Symmetrization InequalityAdvanced
Not assessed3 questions
Contraction InequalityAdvanced
Not assessed1 question
Gemini and Google ModelsFrontier
No quiz
Graph Neural NetworksAdvanced
Not assessed4 questions
Ito's LemmaAdvanced
Not assessed2 questions
Numerical Linear AlgebraFoundations
Not assessed1 question
Peano AxiomsAxioms
Not assessed5 questions
No quiz
Stochastic Calculus for MLAdvanced
Not assessed5 questions
No quiz
Not assessed44 questions
Not assessed3 questions
RLHF and AlignmentResearch
Not assessed3 questions
Transformer ArchitectureResearch
Not assessed11 questions
Sign in to track your mastery and see personalized gap analysis.