Raycaster / evalsBack to SpreadsheetBench

SpreadsheetBench

41348

1/1Pass

SpreadsheetBench task 41348. Inspect the exact spreadsheet prompt, compare published model runs, open the agent response trace, and review grades for workbook editing.

Spreadsheet editingDual harness
ssb-41348
SpreadsheetBench
2 models
Dual (parsed + original)
PromptWorkbook objective and answer cells
ResponseOpen each trace to inspect tool use and edits
GradeCompare score and pass/fail by model

Task prompt

What the agent was asked to do

You are solving a spreadsheet benchmark task in a real workbook. Objective: Produce the correct final workbook state for the expected answer region. What matters: - Only the values in the expected answer region will be graded. - The workbook is the answer. Instructions: 1. Read the workbook and inspect the relevant data region first. 2. Infer the required result for the provided workbook instance. 3. Write the final value(s) directly into the expected answer region. 4. Do not rely on prose, formulas in your chat response, pseudocode, or VBA as the answer unless the benchmark explicitly requires those to be written into cells. 5. If the natural-language task asks for a general method, formula, or macro, convert that into the concrete result needed for this workbook instance. 6. Keep your final text response short and only summarize the workbook cells you changed. Relevant data region(s): Expected answer region(s): Q2:Q11 Expected answer sheet(s): Task: In my Excel file containing project data, each project, identified by a consistent budget ID in column H, may have multiple lines representing different components, such as technology and services listed in column L as revenue types. I need to determine which projects include technology and then mark all associated lines of such projects, regardless of whether a line item itself is technology or service, with a 'Y' in column Q. Projects without a linked budget should be marked as 'N' in column Q. The current formula only marks technology revenue lines and thus doesn't accurately reflect projects that include technology. The sample workbook provided contains  sample projects

Published trajectories

Agent runs on this task

Curated dual-harness runs. Best scored run per model.

ModelHarnessScoreResultLinks
GPT-5.4showcasedual1/1Pass
Gemini 3.1 Produal1/1Pass