Raycaster / evalsBack to SpreadsheetBench

SpreadsheetBench

13284

0/1Fail

SpreadsheetBench task 13284. Inspect the exact spreadsheet prompt, compare published model runs, open the agent response trace, and review grades for workbook editing.

Spreadsheet editingDual harness
ssb-13284
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): 'Base'!E2:E26 Expected answer sheet(s): Task: Return the value from column F in the 'Streets' sheet into column E of the 'Base' sheet based on two match conditions: The first condition is that the value in 'Base'!C must exactly match the value in 'Streets'!B. The second condition is that the numeric value in 'Base'!D must fall within the range defined by the corresponding minimum and maximum values in 'Streets'!C and 'Streets'!D. Only consider rows where both the Start (column C) and End (column D) values are populated. When both conditions are satisfied, return the matching name from 'Streets'!F. If no match is found, return an empty string.

Published trajectories

Agent runs on this task

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

ModelHarnessScoreResultLinks
GPT-5.4showcasedual0/1Fail
Gemini 3.1 Produal0/1Fail