Lab interactive

what you're seeing

try this

1. 1.Press Run analysis and watch which nodes light up together.
2. 2.Run it a few times — when the rating comes out Sell while the weight is high, the divergence warning appears.
3. 3.Raise the speed to see the shape of the graph instead of the text.

real vs simplified

what you're seeing

try this

1. 1.Start at 9B · Q4_K_M · 8 GB laptop with 32k context: it doesn't fit while the display uses the GPU.
2. 2.Turn on “move display to iGPU” — the last layers come back to the GPU.
3. 3.Switch to 27B · AWQ on 24 GB, then on 48 GB, and shrink the context to 16k.

real vs simplified

what you're seeing

try this

1. 1.Watch an invariant fail in the middle and the agent fix it using a note from memory.
2. 2.At the end, approve or reject — rejection sends the plan back.
3. 3.Replay at higher speed to see the layers alternate.

real vs simplified

what you're seeing

try this

1. 1.Turn sparse sampling off and on — without it, the cloud never reaches the upper edge.
2. 2.Switch from a dovish to a hawkish regime and watch crypto and equities shrink.
3. 3.Lower the number of simulations to see how noisy the optimum gets.

real vs simplified

what you're seeing

try this

1. 1.Push the bias to dovish, then raise the rate above 10%: the override reads the cut as normalization (neutral).
2. 2.Pick each event and compare the raw reaction with the final adjustment.
3. 3.Notice the clip: a big reaction never moves expected returns by more than the cap.

real vs simplified

what you're seeing

try this

1. 1.Click Mutate legacy, then Run sync: only the amber months travel.
2. 2.Switch to “all at once” and run again — memory blows past the limit.
3. 3.In the daily rebuild, turn on “chain dies silently”: the dead chain stays visible as pending work.

real vs simplified

what you're seeing

try this

1. 1.Tab A: move the rebuild into the amber window and run it — two columns turn to zero and the log still says OK. Then restore from _stage.
2. 2.Tab B: watch which hours abort, then turn on the fixes one at a time.
3. 3.Compare the abort rate between “guard ok” and “guard gave up” in the correlation table.

real vs simplified

what you're seeing

try this

1. 1.Click a source to trace its path to the dashboards.
2. 2.Filter by RULER and CONFERENCE — they're never ingested like a pipeline.
3. 3.Click “manual source goes stale” and see which dashboards inherit the warning.

real vs simplified

what you're seeing

try this

1. 1.Go from Landing to Silver and spot the duplicate order disappearing and the timestamp moving to UTC.
2. 2.In Gold, switch USD/BRL: every money column exists in both currencies.
3. 3.Read the partition-pruning note — it's the most common reason a mart query scans everything.

real vs simplified

what you're seeing

try this

1. 1.Start in proportional mode: a day with spend but no views loses its cost.
2. 2.Switch to exact, then turn on “union of all landing-dates” — the lost cost comes back.
3. 3.Turn on “collapse per player” to stop a shared player's revenue from counting twice. Then add a new creative.

real vs simplified

what you're seeing

try this

1. 1.Click “REVOKE FROM anon” — anonymous still gets 200. That's the PUBLIC trap.
2. 2.Now revoke from PUBLIC and anon: anonymous is out, but the member still reads master-only rows.
3. 3.Turn on the internal role guard: each role sees only its tier.

real vs simplified

what you're seeing

try this

1. 1.Switch theme: surfaces change, data colors don't.
2. 2.View as Member: order counts and AOV disappear — from the table and from the CSV.
3. 3.Change the period and check that the CSV preview always matches the table.

real vs simplified

what you're seeing

try this

1. 1.Switch to “net with refunds so far” and set the sale age to 7 days — the margin looks much better than it will be.
2. 2.Drag the age toward 180 days and watch the gap close.
3. 3.Click a row of the factors table to highlight where it enters the waterfall.

real vs simplified

what you're seeing

try this

1. 1.Pause SMS: dots pile up as stuck and the match rate drops — then resume and they're delivered late.
2. 2.Switch Events ⇄ Leads to see how much the unit changes the match rate.
3. 3.Tap a dot to open that lead's Gantt; check the matrix — unrouted pairs never count as stuck.

real vs simplified

what you're seeing

try this

1. 1.As Member, hide only AOV: the detector shows it leaks through revenue ÷ orders.
2. 2.Hide order count too — now the cut is consistent.
3. 3.Turn on CSV export: the file carries the hidden columns until “CSV follows the screen cut” is on.

real vs simplified