# Pedagogical Knowledge Evaluation on CDPK (Cross-Domain Pedagogical Knowledge) Benchmark Source: [Pedagogical Knowledge Evaluation on CDPK (Cross-Domain Pedagogical Knowledge) Benchmark]() Collection index: [llms.txt]() Data updated: 2026-04-09T23:18:13.616Z Evaluation of model performance on the Cross-Domain Pedagogical Knowledge (CDPK) Benchmark, focusing on pedagogical knowledge assessment using metrics like Replicated Accuracy and Reported 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 | Pedagogical Knowledge Evaluation | | Dataset (including any stated split) | CDPK (Cross-Domain Pedagogical Knowledge) Benchmark | | Metric | Replicated Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 7 of 7 | | 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 - [Replicated Accuracy](): Higher is better; unit: PERCENT. - [Reported Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Replicated Accuracy | Result date | Report paper | | --- | --- | --- | --- | | EDUQWEN 32B-SFT-RL2 | 96.52 | Not recorded | 69d732717d0172140ec76a15 | | EDUQWEN 32B-SFT | 96.2 | Not recorded | 69d732717d0172140ec76a15 | | EDUQWEN 32B-RL1 | 94.13 | Not recorded | 69d732717d0172140ec76a15 | | QWEN3-235B-A22B-THINKING | 86.63 | Not recorded | 69d732717d0172140ec76a15 | | QWEN3-32B BASE MODEL | 83.37 | Not recorded | 69d732717d0172140ec76a15 | | QWEN2.5-72B-INSTRUCT | 78.6 | Not recorded | 69d732717d0172140ec76a15 | ### EDUQWEN 32B-SFT-RL2 Replicated Accuracy: 96.52. Result ID: 69d833a647f34ce924bfe1a8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Inference-time compute":"RL 2ND ROUND"}. This record does not assert independent reproduction. ### EDUQWEN 32B-SFT Replicated Accuracy: 96.2. Result ID: 69d833a647f34ce924bfe1a7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Inference-time compute":"SFT"}. This record does not assert independent reproduction. ### EDUQWEN 32B-RL1 Replicated Accuracy: 94.13. Result ID: 69d833a647f34ce924bfe1a6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Inference-time compute":"RL 1ST ROUND"}. This record does not assert independent reproduction. ### QWEN3-235B-A22B-THINKING Replicated Accuracy: 86.63. Result ID: 69d833a647f34ce924bfe1a3. [Reporting paper]() Recorded evaluation configuration: {"Inference-time compute":"RANDOM FEW-SHOT"}. This record does not assert independent reproduction. ### QWEN3-32B BASE MODEL Replicated Accuracy: 83.37. Result ID: 69d833a647f34ce924bfe1a5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Inference-time compute":"RANDOM FEW-SHOT"}. This record does not assert independent reproduction. ### QWEN2.5-72B-INSTRUCT Replicated Accuracy: 78.6. Result ID: 69d833a647f34ce924bfe1a4. [Reporting paper]() Recorded evaluation configuration: {"Inference-time compute":"RANDOM FEW-SHOT"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.