The MIT AI Lab three questions (Q1/Q2/Q3) activate deep reading when combined with Keshav Three-Pass as the PUMA research protocol

Reading a research paper well requires two distinct cognitive modes: systematic coverage (knowing what to look at and in what order) and generative engagement (actively thinking while reading). The Keshav Three-Pass Method provides the former; MIT AI Lab Working Paper 316 provides the latter.


What WP 316 Actually Says

MIT AI Lab Working Paper 316 — “How to Do Research at the MIT AI Lab” (1988, ed. David Chapman) — is an informal internal guide, not a peer-reviewed publication. It mentions three reading questions as habits of mind:

Q1 — “How can I use this?” Map the paper’s methods, metrics, datasets, or framings to your own research. This prevents passive absorption: every paper becomes a candidate resource for your own work.

Q2 — “Does this really do what the author claims?” Active scepticism. Evaluate whether the evaluation is rigorous, whether baselines are fair, whether the claim is over-stated relative to the evidence.

Q3 — “What if…?” Generative extension. Relax an assumption. Apply the method to a different domain. Think about what would break. This is the seed of research contributions.

These questions are informal suggestions in WP 316, not a formalised methodology. They are most useful during deep reading (Keshav Pass 3 equivalent).


The Integrated PUMA Reading Protocol

For PUMA’s SLR (≥40 papers, OE1), the reading protocol combines both frameworks:

PAPER ENCOUNTERED
      ↓
KESHAV PASS 1 (5–10 min)
   → Read: title, abstract, intro, headings, conclusions, references
   → Evaluate 5Cs: Category · Context · Correctness · Contributions · Clarity
   → Decision: PASS 2 or ARCHIVE
      ↓ (if relevant)
KESHAV PASS 2 (≤1 hour)
   → Read carefully, skip proofs
   → Note key points, mark unread references
   → Fleeting note → Literature note in [20 - Literature/]
      ↓ (for core papers only)
KESHAV PASS 3 + WP316 Q1/Q2/Q3 (2–5 hours)
   → Q1: "How can I use this?" → map to PUMA tasks/metrics/datasets
   → Q2: "Does this really do what the author claims?" → scrutinise baselines, evaluation
   → Q3: "What if…?" → generate extensions, challenges, new hypotheses
   → Virtual re-implementation of core contribution
   → Permanent note in [30 - Permanent/] with declarative title

AI-Assisted Application (CDD + RCOIF)

Each WP 316 question can be operationalised with a targeted AI prompt using RCOIF:

Q1 prompt template:

ROLE: Expert researcher in LLM benchmarking for PM tasks.
CONTEXT: I am building PUMA, a reproducible local-LLM benchmark for ICT project management.
OBJECTIVE: Extract from the following paper abstract all methods, metrics, datasets, and framings directly applicable to PUMA.
INSTRUCTIONS: For each applicable element, state: what it is, how it applies, and what would need to be adapted.
FORMAT: Bullet list grouped by: Metrics | Methods | Datasets | Framings.
[Paste paper abstract]

Q2 prompt template:

ROLE: Statistical critic reviewing a paper for a peer-reviewed venue.
CONTEXT: [Paste paper claim + methods section excerpt]
OBJECTIVE: Identify all validity threats: are baselines appropriate? Is the evaluation fair? Is the claim over-stated?
INSTRUCTIONS: Apply AMI — diagnose weaknesses first, then suggest how the authors could have strengthened the study.
FORMAT: (1) Weaknesses list · (2) Suggested improvements · (3) PUMA implication.

Q3 prompt template:

ROLE: Creative research thinker specialising in LLM evaluation.
CONTEXT: [Paste paper core contribution in 2 sentences]
OBJECTIVE: Generate 5 "What if…?" extensions that a researcher could investigate.
INSTRUCTIONS: Each extension must be: (a) falsifiable, (b) relevant to PM+LLM, (c) distinct from what the paper already tests.
FORMAT: Numbered list with: Extension · Why interesting · Connection to PUMA.

The “MIT Student Method” Label

Various Spanish-language academic guides label the combination of WP 316’s three questions + Keshav Three-Pass + AI-prompting frameworks (RCOIF, EGI, AMI, DRCA, IIPR) as the “MIT Student Method.” This label is a pedagogical shorthand, not an official MIT publication. For academic rigour, PUMA cites:

  • WP 316 directly for the three questions
  • Keshav (2007) for the Three-Pass protocol
  • Individual references for each prompting framework

Relationship to Other PUMA Frameworks

FrameworkRoleWhen to use
Keshav Three-PassSystematic reading protocolEvery paper in SLR
WP 316 Q1/Q2/Q3Active engagement mindsetDuring Pass 3 of core papers
RCOIFAI prompt structureAny AI-assisted question
EGIExploratory AI dialogueUnfamiliar territory / new topic
AMIIterative AI self-critiqueImproving writing or prompts
DRCADeep concept deconstructionComplex methodological papers
IIPRReverse-engineer AI failuresWhen prompts underperform

References

  • MIT AI Lab. (1988). How to do research at the MIT AI Lab (AI Lab Working Paper 316). MIT.
  • Keshav, S. (2007). How to read a paper. ACM SIGCOMM Computer Communication Review, 37(3), 83–84.
  • LN-MITAILab-WP316-HowToDoResearch

MOCs