# Knowledge-to-Action QA on 50 Knowledge-to-Action Questions (gold-standard) Source: [Knowledge-to-Action QA on 50 Knowledge-to-Action Questions (gold-standard)]() Collection index: [llms.txt]() Data updated: 2026-04-10T06:58:34.068Z Evaluation of Knowledge-to-Action Question Answering performance using the 50 knowledge-to-action questions gold-standard dataset, measured using ROUGE metrics. 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 | Knowledge-to-action question answering | | Dataset (including any stated split) | 50 knowledge-to-action questions gold-standard | | Metric | ROUGE-1 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 9 of 9 | | 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 - [ROUGE-1](): Higher is better; unit: PERCENT. - [ROUGE-L](): Higher is better; unit: PERCENT. ## Recorded results | Method | ROUGE-1 | Result date | Report paper | | --- | --- | --- | --- | | vanilla LLM | 29.9 | Not recorded | 68fa7bef1e209131cea8b6da | | text-chunk RAG | 29.6 | Not recorded | 68fa7bef1e209131cea8b6da | | vanilla LLM | 29.3 | Not recorded | 68fa7bef1e209131cea8b6da | | KEO | 29.3 | Not recorded | 68fa7bef1e209131cea8b6da | | KEO | 29.2 | Not recorded | 68fa7bef1e209131cea8b6da | | text-chunk RAG | 21.3 | Not recorded | 68fa7bef1e209131cea8b6da | | KEO | 12.4 | Not recorded | 68fa7bef1e209131cea8b6da | | vanilla LLM | 11 | Not recorded | 68fa7bef1e209131cea8b6da | | text-chunk RAG | 10.7 | Not recorded | 68fa7bef1e209131cea8b6da | ### vanilla LLM ROUGE-1: 29.9. Result ID: 69d89e1c47f34ce924c0ec1e. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"mistral-nemo-it"}. This record does not assert independent reproduction. ### text-chunk RAG ROUGE-1: 29.6. Result ID: 69d89e1c47f34ce924c0ec17. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"gemma-3-it"}. This record does not assert independent reproduction. ### vanilla LLM ROUGE-1: 29.3. Result ID: 69d89e1c47f34ce924c0ec18. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"gemma-3-it"}. This record does not assert independent reproduction. ### KEO ROUGE-1: 29.3. Result ID: 69d89e1c47f34ce924c0ec19. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"gemma-3-it"}. This record does not assert independent reproduction. ### KEO ROUGE-1: 29.2. Result ID: 69d89e1c47f34ce924c0ec1f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"mistral-nemo-it"}. This record does not assert independent reproduction. ### text-chunk RAG ROUGE-1: 21.3. Result ID: 69d89e1c47f34ce924c0ec1d. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"mistral-nemo-it"}. This record does not assert independent reproduction. ### KEO ROUGE-1: 12.4. Result ID: 69d89e1c47f34ce924c0ec1c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"phi-4"}. This record does not assert independent reproduction. ### vanilla LLM ROUGE-1: 11. Result ID: 69d89e1c47f34ce924c0ec1b. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"phi-4"}. This record does not assert independent reproduction. ### text-chunk RAG ROUGE-1: 10.7. Result ID: 69d89e1c47f34ce924c0ec1a. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"phi-4"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.