# Spreadsheet Reasoning on SpreadsheetBench Verified Source: [Spreadsheet Reasoning on SpreadsheetBench Verified]() Collection index: [llms.txt]() Data updated: 2026-05-28T07:54:30.353Z Evaluation of model performance on the SpreadsheetBench Verified dataset for the Spreadsheet Reasoning task, measured using accuracy. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Spreadsheet Reasoning | | Dataset (including any stated split) | SpreadsheetBench Verified | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | SkillGrad | 71.11 | Not recorded | 6a17cc27dbd7150506f8139c | | SkillGrad | 69.44 | Not recorded | 6a17cc27dbd7150506f8139c | | EvoSkill | 68.06 | Not recorded | 6a17cc27dbd7150506f8139c | | Trace2Skill | 65.28 | Not recorded | 6a17cc27dbd7150506f8139c | | Trace2Skill | 63.89 | Not recorded | 6a17cc27dbd7150506f8139c | | EvoSkill | 63.61 | Not recorded | 6a17cc27dbd7150506f8139c | | No Skill | 62.5 | Not recorded | 6a17cc27dbd7150506f8139c | | Base xlsx Skill | 60 | Not recorded | 6a17cc27dbd7150506f8139c | | Base xlsx Skill | 55.83 | Not recorded | 6a17cc27dbd7150506f8139c | | SkillGrad | 54.17 | Not recorded | 6a17cc27dbd7150506f8139c | | SkillGrad | 45.83 | Not recorded | 6a17cc27dbd7150506f8139c | | No Skill | 44.17 | Not recorded | 6a17cc27dbd7150506f8139c | | Trace2Skill | 38.89 | Not recorded | 6a17cc27dbd7150506f8139c | | Trace2Skill | 37.22 | Not recorded | 6a17cc27dbd7150506f8139c | | EvoSkill | 37.22 | Not recorded | 6a17cc27dbd7150506f8139c | | EvoSkill | 36.94 | Not recorded | 6a17cc27dbd7150506f8139c | | Base xlsx Skill | 36.67 | Not recorded | 6a17cc27dbd7150506f8139c | | Base xlsx Skill | 33.33 | Not recorded | 6a17cc27dbd7150506f8139c | ### SkillGrad Accuracy: 71.11. Result ID: 6a17f4a2fbc373592c117a2a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"LLM-gen.","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### SkillGrad Accuracy: 69.44. Result ID: 6a17f4a2fbc373592c117a2d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"Third-party","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### EvoSkill Accuracy: 68.06. Result ID: 6a17f4a2fbc373592c117a29. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"LLM-gen.","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### Trace2Skill Accuracy: 65.28. Result ID: 6a17f4a2fbc373592c117a28. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"LLM-gen.","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### Trace2Skill Accuracy: 63.89. Result ID: 6a17f4a2fbc373592c117a2b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"Third-party","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### EvoSkill Accuracy: 63.61. Result ID: 6a17f4a2fbc373592c117a2c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"Third-party","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### No Skill Accuracy: 62.5. Result ID: 6a17f4a2fbc373592c117a25. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"None","Training Type":"Training-free"}. This record does not assert independent reproduction. ### Base xlsx Skill Accuracy: 60. Result ID: 6a17f4a2fbc373592c117a27. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"Third-party","Training Type":"Training-free"}. This record does not assert independent reproduction. ### Base xlsx Skill Accuracy: 55.83. Result ID: 6a17f4a2fbc373592c117a26. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GPT-5.4","Initialization":"LLM-gen.","Training Type":"Training-free"}. This record does not assert independent reproduction. ### SkillGrad Accuracy: 54.17. Result ID: 6a17f4a2fbc373592c117a33. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"LLM-gen.","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### SkillGrad Accuracy: 45.83. Result ID: 6a17f4a2fbc373592c117a36. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"Third-party","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### No Skill Accuracy: 44.17. Result ID: 6a17f4a2fbc373592c117a2e. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"None","Training Type":"Training-free"}. This record does not assert independent reproduction. ### Trace2Skill Accuracy: 38.89. Result ID: 6a17f4a2fbc373592c117a34. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"Third-party","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### Trace2Skill Accuracy: 37.22. Result ID: 6a17f4a2fbc373592c117a31. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"LLM-gen.","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### EvoSkill Accuracy: 37.22. Result ID: 6a17f4a2fbc373592c117a32. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"LLM-gen.","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### EvoSkill Accuracy: 36.94. Result ID: 6a17f4a2fbc373592c117a35. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"Third-party","Training Type":"Training-based skill improvement"}. This record does not assert independent reproduction. ### Base xlsx Skill Accuracy: 36.67. Result ID: 6a17f4a2fbc373592c117a2f. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"LLM-gen.","Training Type":"Training-free"}. This record does not assert independent reproduction. ### Base xlsx Skill Accuracy: 33.33. Result ID: 6a17f4a2fbc373592c117a30. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"GPT-4.1","Initialization":"Third-party","Training Type":"Training-free"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.