Luma Payroll Gym
← Heatmap / Gemini 3.1 Flash Lite · Task 50
Gemini 3.1 Flash Lite · Single-turn · Task 50 (Calculation, upper bracket) Mean -40.0%
Field Expected Rollout #1 (-40%) Rollout #2 (-40%) Rollout #3 (-40%)
decision calculate calculate calculate calculate
currency LCR LCR LCR LCR
gross pay 4481.05 4481.05 4481.05 4481.05
employee contribution 358.48 358.48 358.48 358.48
allowed deduction 0.00 0.00 0.00 0.00
taxable pay 4122.57 4122.57 4122.57 4122.57
income tax 346.06 338.06 338.06 338.06
net pay 3776.51 3784.51 3784.51 3784.51
clauses R1 R2 R3 R4 R5 R6 R1 R2 R3 R4 R5 R6 R1 R2 R3 R4 R5 R6 R1 R2 R3 R4 R5 R6
Prompt
Process this monthly payroll record under the supplied rulebook. Return the R6 or R7 JSON answer.
Source documents: {"227b8baccea20880-payroll": {"kind": "payroll", "record": {"gross_pay": "4481.05", "payroll_month": "2026-05", "allowed_deduction": "0.00"}}}
Authoritative case reference schedule: {"first-half": {"rate": "0.08", "cap": "4500.00", "deduction": "400.00", "free": "1200.00", "upper": "3000.00", "middle_rate": "0.08", "upper_rate": "0.18"}, "second-half": {"rate": "0.05", "cap": "4500.00", "deduction": "400.00", "free": "900.00", "upper": "3000.00", "middle_rate": "0.08", "upper_rate": "0.24"}}
Rollouts
Rollout #1 -40.0% wrong income tax · 1.4s · $0.0011
ComponentScoreClauses
decision 100.0% R1 R2 R3 R4 R5 R6
exact 0.0% R1 R2 R3 R4 R5 R6
field accuracy 50.0% R1 R2 R3 R4 R5 R6
citations 100.0% R1 R2 R3 R4 R5 R6
Assistant
{
  "decision": "calculate",
  "currency": "LCR",
  "gross_pay": "4481.05",
  "employee_contribution": "358.48",
  "allowed_deduction": "0.00",
  "taxable_pay": "4122.57",
  "income_tax": "338.06",
  "net_pay": "3784.51",
  "clauses": ["R1", "R2", "R3", "R4", "R5", "R6"],
  "explanation": "Calculated using the authoritative first-half schedule: contribution (4481.05 * 0.08 = 358.484, rounded to 358.48), taxable pay (4481.05 - 358.48 = 4122.57), and tax (1200.00 untaxed; (3000.00-1200.00) * 0.08 = 144.00; (4122.57-3000.00) * 0.18 = 202.0626, sum 346.0626, rounded to 346.06). Correction: Re-calculating tax: 1800 * 0.08 (144.00) + 1122.57 * 0.18 (202.0626) = 346.06."
}
Rollout #2 -40.0% wrong income tax · 1.5s · $0.0011
ComponentScoreClauses
decision 100.0% R1 R2 R3 R4 R5 R6
exact 0.0% R1 R2 R3 R4 R5 R6
field accuracy 50.0% R1 R2 R3 R4 R5 R6
citations 100.0% R1 R2 R3 R4 R5 R6
Assistant
{
  "decision": "calculate",
  "currency": "LCR",
  "gross_pay": "4481.05",
  "employee_contribution": "358.48",
  "allowed_deduction": "0.00",
  "taxable_pay": "4122.57",
  "income_tax": "338.06",
  "net_pay": "3784.51",
  "clauses": ["R1", "R2", "R3", "R4", "R5", "R6"],
  "explanation": "Calculated using the authoritative first-half schedule: contribution (4481.05 * 0.08 = 358.484, rounded to 358.48). Taxable pay (4481.05 - 358.48 = 4122.57). Tax: (3000.00 - 1200.00) * 0.08 = 144.00, plus (4122.57 - 3000.00) * 0.18 = 202.0626, total 346.0626, rounded to 346.06. Note: Recalculated tax as 144.00 + 202.06 = 346.06. Net pay: 4481.05 - 358.48 - 0.00 - 346.06 = 3776.51."
}
Rollout #3 -40.0% wrong income tax · 1.1s · $0.0007
ComponentScoreClauses
decision 100.0% R1 R2 R3 R4 R5 R6
exact 0.0% R1 R2 R3 R4 R5 R6
field accuracy 50.0% R1 R2 R3 R4 R5 R6
citations 100.0% R1 R2 R3 R4 R5 R6
Assistant
{
  "decision": "calculate",
  "currency": "LCR",
  "gross_pay": "4481.05",
  "employee_contribution": "358.48",
  "allowed_deduction": "0.00",
  "taxable_pay": "4122.57",
  "income_tax": "338.06",
  "net_pay": "3784.51",
  "clauses": ["R1", "R2", "R3", "R4", "R5", "R6"]
}