# Agentic Instruction Following on TaskCraft Source: [Agentic Instruction Following on TaskCraft]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:57:50.945Z Evaluation of model performance on agentic instruction following capabilities using the TaskCraft dataset. 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 | Agentic Instruction Following | | Dataset (including any stated split) | TaskCraft | | Metric | Performance Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | 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 - [Performance Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Performance Score | Result date | Report paper | | --- | --- | --- | --- | | MemEvolve + Flash-Searcher | 79.33 | Not recorded | 694a223d9eebe296e72aa99d | | MemEvolve + SmolAgent | 77 | Not recorded | 694a223d9eebe296e72aa99d | | Agent KB | 75.33 | Not recorded | 694a223d9eebe296e72aa99d | | MemEvolve + SmolAgent | 75 | Not recorded | 694a223d9eebe296e72aa99d | | MemEvolve + Flash-Searcher | 75 | Not recorded | 694a223d9eebe296e72aa99d | | Agent KB | 72.67 | Not recorded | 694a223d9eebe296e72aa99d | | MemEvolve + Flash-Searcher | 72.67 | Not recorded | 694a223d9eebe296e72aa99d | | MemEvolve + Flash-Searcher | 72 | Not recorded | 694a223d9eebe296e72aa99d | | Flash-Searcher | 69.67 | Not recorded | 694a223d9eebe296e72aa99d | | Flash-Searcher | 69.33 | Not recorded | 694a223d9eebe296e72aa99d | | MemEvolve + Flash-Searcher | 68 | Not recorded | 694a223d9eebe296e72aa99d | | MemEvolve + SmolAgent | 67.67 | Not recorded | 694a223d9eebe296e72aa99d | | Cognitive Kernel-Pro | 66 | Not recorded | 694a223d9eebe296e72aa99d | | Smolagents | 64 | Not recorded | 694a223d9eebe296e72aa99d | | Agent KB | 61.67 | Not recorded | 694a223d9eebe296e72aa99d | | OWL Workforce | 58.33 | Not recorded | 694a223d9eebe296e72aa99d | | Flash-Searcher | 58 | Not recorded | 694a223d9eebe296e72aa99d | ### MemEvolve + Flash-Searcher Performance Score: 79.33. Result ID: 6996b68a5698665ec40117f9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"GPT-5-mini","pass@k":"3"}. This record does not assert independent reproduction. ### MemEvolve + SmolAgent Performance Score: 77. Result ID: 6996b68a5698665ec40117f6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"GPT-5-mini","pass@k":"3"}. This record does not assert independent reproduction. ### Agent KB Performance Score: 75.33. Result ID: 6996b68a5698665ec40117f0. [Reporting paper]() Recorded evaluation configuration: {"Model Family":"GPT-4.1","pass@k":"3"}. This record does not assert independent reproduction. ### MemEvolve + SmolAgent Performance Score: 75. Result ID: 6996b68a5698665ec40117f5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"GPT-5-mini","pass@k":"2"}. This record does not assert independent reproduction. ### MemEvolve + Flash-Searcher Performance Score: 75. Result ID: 6996b68a5698665ec40117f8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"GPT-5-mini","pass@k":"2"}. This record does not assert independent reproduction. ### Agent KB Performance Score: 72.67. Result ID: 6996b68a5698665ec40117ef. [Reporting paper]() Recorded evaluation configuration: {"Model Family":"GPT-4.1","pass@k":"2"}. This record does not assert independent reproduction. ### MemEvolve + Flash-Searcher Performance Score: 72.67. Result ID: 6996b68a5698665ec40117fb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"DeepSeek V3.2","pass@k":"1"}. This record does not assert independent reproduction. ### MemEvolve + Flash-Searcher Performance Score: 72. Result ID: 6996b68a5698665ec40117f7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"GPT-5-mini","pass@k":"1"}. This record does not assert independent reproduction. ### Flash-Searcher Performance Score: 69.67. Result ID: 6996b68a5698665ec40117f1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"GPT-5-mini","pass@k":"1"}. This record does not assert independent reproduction. ### Flash-Searcher Performance Score: 69.33. Result ID: 6996b68a5698665ec40117f3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"DeepSeek V3.2","pass@k":"1"}. This record does not assert independent reproduction. ### MemEvolve + Flash-Searcher Performance Score: 68. Result ID: 6996b68a5698665ec40117fa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"Kimi K2","pass@k":"1"}. This record does not assert independent reproduction. ### MemEvolve + SmolAgent Performance Score: 67.67. Result ID: 6996b68a5698665ec40117f4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"GPT-5-mini","pass@k":"1"}. This record does not assert independent reproduction. ### Cognitive Kernel-Pro Performance Score: 66. Result ID: 6996b68a5698665ec40117ed. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"Claude-3.7 etc.","pass@k":"1"}. This record does not assert independent reproduction. ### Smolagents Performance Score: 64. Result ID: 6996b68a5698665ec40117ec. [Reporting paper]() Recorded evaluation configuration: {"Model Family":"GPT-5-mini"}. This record does not assert independent reproduction. ### Agent KB Performance Score: 61.67. Result ID: 6996b68a5698665ec40117ee. [Reporting paper]() Recorded evaluation configuration: {"Model Family":"GPT-4.1","pass@k":"1"}. This record does not assert independent reproduction. ### OWL Workforce Performance Score: 58.33. Result ID: 6996b68a5698665ec40117eb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"GPT-4o+o3-mini","pass@k":"3"}. This record does not assert independent reproduction. ### Flash-Searcher Performance Score: 58. Result ID: 6996b68a5698665ec40117f2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Family":"Kimi K2","pass@k":"1"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.