Match example difficulty to growing competence.

Curriculum learning changes the order or sampling weight of training examples. Compare schedules while watching coverage, loss, competence, and forgetting evolve.

Run the curriculum

Difficulty admitted over training

Eligible difficulty ≤ competence 0.20
Competence0.20

Current modeled ability.

Dataset coverage20%

Examples eligible under the pace.

Training loss1.00

Challenge adjusted by current skill.

Forgetting risk0%

Penalty when easy examples disappear.

Ordering is an optimization choice.

Pacing controls exposure

A curriculum defines when examples become eligible. Too slow wastes data; too fast behaves like uniform sampling.

Difficulty needs a signal

Human rules, model loss, confidence, length, noise, or task structure can rank examples, but every proxy can encode bias.

Revisit earlier material

Hard-first transitions can destabilize learning, while removing easy examples entirely can cause forgetting. Mix replay with expansion.

Primary reading

Curriculum Learning Bengio et al.Automated Curriculum Learning for Neural Networks Graves et al.Curriculum Learning: A Survey Wang et al.