← Glean-ia-acs · Day 27 of 56
Level 300 — Power user30 minNIST MAP

Assistant Thinking Mode

When to invoke multi-step reasoning

You'll be able to

Read first

Day 6 Lab

  1. Write three questions: one lookup, one synthesis, one comparison across sources.
  2. Predict which would benefit from a multi-step reasoning pass and which would not.
  3. Run each and observe answer depth versus the time it took to return.
  4. Note where the deeper pass changed the answer, not just its length.
  5. Draft a personal rule for when to reach for thinking mode.

Working Example: Matching the Question to the Mode

A multi-step "thinking" mode is among the capabilities Glean described at its Fall '25 launch; attribute the capability to Glean rather than asserting it from first principles. The skill is matching the question to the mode. Verified against fall25_launch.

Question shape                          Better mode (per Glean's Fall '25 framing)
-------------------------------------   -----------------------------------------
"What's our PTO accrual rate?"          quick answer (single fact lookup)
"Compare our three vendor SLAs and      multi-step reasoning (synthesis across
 flag the weakest term"                  several docs, then a judgment)
"Summarize this one doc"                quick answer (one source, no cross-ref)

Worth-the-wait test:
  Does answering require pulling from SEVERAL sources and reasoning over them?
    yes → thinking mode earns its latency
    no  → a quick answer is faster and just as correct

Check for understanding

Give one question where a deeper multi-step pass is worth the wait, and one where a quick answer is better.

Check yourself

Pick an answer — you'll see if it's right and why.

1. When stacking operators on a 500-result query, how do you know which constraint did the real work?

2. You're exploring an unfamiliar result set and don't yet know which apps have hits. What's the better tool?

3. A trusted document ranks at #8. What's the honest way to investigate the ranking?

4. What three things must a written search recipe include so a teammate reproduces your exact result set?

5. Which question best justifies Glean's multi-step thinking mode over a quick answer?

Pass threshold 80% · week 4 quiz

In 10 seconds

“Not every question needs deep reasoning. Glean's Fall '25 launch describes a thinking mode for multi-step work — use it for synthesis and comparison, skip it for a single lookup.”

Next · Week 4 Day 7
Prompts That Route Reliably
← W04D05all lessonsW04D07 →