Skip to content

Latest commit

 

History

History
52 lines (36 loc) · 3.03 KB

File metadata and controls

52 lines (36 loc) · 3.03 KB

Inside MiniMax course brief

Status: accepted concept; source freeze required before implementation Prerequisites: AI from First Principles v0.2 and Open Frontier Models: Shared Techniques v0.1 Candidate release: MiniMax M1

Promise

Reconstruct selected hybrid-attention, long-context, and reasoning techniques from a pinned MiniMax release and evaluate their trade-offs on deterministic sequence tasks.

Learning outcomes

After completing the course, a learner can:

  • read the pinned release's architecture and evidence package critically;
  • implement toy full and efficient attention paths under one interface;
  • build and test a hybrid layer schedule;
  • compare memory, work, state growth, and recall as sequence length changes;
  • identify adversarial sequences that expose efficient-attention limitations;
  • evaluate a reasoning-budget claim separately from long-context behavior.

Provisional curriculum

  1. Reading MiniMax M1 — release artifacts, configuration, licences, attention schedule, reported claims, and missing evidence.
  2. Efficient attention state — recurrent update, decay, retrieval, numerical behavior, and causal invariants.
  3. Hybrid attention — layer schedules, local/global information paths, memory, compute, and failure cases.
  4. Long-context experiments — retrieval distance, distractors, state capacity, cost scaling, and claim-aligned metrics.
  5. Reasoning and context — sampled work, budgets, stopping, reward evidence, and avoiding causal stories unsupported by the experiment.

Explorable and exercise spine

Lesson Explorable Focused implementation
Release anatomy Evidence and layer-schedule map Parse the pinned architecture
Efficient state Recurrent state stepper Implement a stable state update
Hybrid attention Edit the layer schedule and trace information Build a valid hybrid schedule
Long context Cost/recall curves with distractors Implement the retrieval benchmark
Reasoning Context/budget/score comparison Validate a claim-aligned protocol

Capstone

Replace selected foundation attention layers with a toy efficient path while retaining periodic full attention. Compare at least two schedules on recall, memory, compute, and adversarial sequences. State why the result does not reproduce production long-context or reasoning capability.

Boundary

This course owns the pinned MiniMax interpretation and hybrid composition. Generic experiment design and cost units come from the shared core; efficient attention learned in another model course is optional context, not a hidden prerequisite.

Acceptance criteria

  • The exact report, repository commit, model card, weight terms, and evaluation protocol are frozen.
  • Efficient and full paths have causal and numerical unit tests.
  • Hybrid schedules are deterministic and preserve declared layer counts.
  • Long-context results report both quality and resource costs.
  • The course passes clean-checkout validation without production weights, accounts, APIs, GPUs, or network services.