# Misalignment Detection on Workbench n=62 (general) Source: [Misalignment Detection on Workbench n=62 (general)]() Collection index: [llms.txt]() Data updated: 2026-07-03T06:22:16.339Z Evaluation of Misalignment Detection performance on the Workbench n=62 (general) dataset, reporting Error Rate (ER) and p-value. 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 | Misalignment Detection | | Dataset (including any stated split) | Workbench n=62 (general) | | Metric | Error Rate (ER) | | Unit | PERCENT | | Direction | Lower is better | | Results loaded | 30 of 30 | | 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 - [Error Rate (ER)](): Lower is better; unit: PERCENT. - [Delta](): Lower is better; unit: SCALAR. - [p-value](): Lower is better; unit: SCALAR. ## Recorded results | Method | Error Rate (ER) | Result date | Report paper | | --- | --- | --- | --- | | ProvenanceGuard | 0.3 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 1.6 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 1.9 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 2.3 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 2.9 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 3.5 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 5.6 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 7.4 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 9.7 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 10.3 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 10.3 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 12.9 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 13.9 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 14.5 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 15.2 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 15.2 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 15.5 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 16.8 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 19.4 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 26.5 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 29 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 32.9 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 34.8 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 37.4 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 47.7 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 49 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvenanceGuard | 51 | Not recorded | 6a4741ee143ebdcacacc5a79 | | ProvePrompt | 59.7 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 64.8 | Not recorded | 6a4741ee143ebdcacacc5a79 | | DirectPrompt | 66.8 | Not recorded | 6a4741ee143ebdcacacc5a79 | ### ProvenanceGuard Error Rate (ER): 0.3. Result ID: 6a4754fa60b88bb33991b281. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Gemini 3 Flash"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 1.6. Result ID: 6a4754fa60b88bb33991b282. [Reporting paper]() Recorded evaluation configuration: {"Model":"Gemini 3 Flash"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 1.9. Result ID: 6a4754fa60b88bb33991b276. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-5"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 2.3. Result ID: 6a4754fa60b88bb33991b277. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"GPT-5"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 2.9. Result ID: 6a4754fa60b88bb33991b278. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-5 Mini"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 3.5. Result ID: 6a4754fa60b88bb33991b283. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Gemini 3 Flash"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 5.6. Result ID: 6a4754fa60b88bb33991b290. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Claude Haiku 4.5"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 7.4. Result ID: 6a4754fa60b88bb33991b275. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-5"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 9.7. Result ID: 6a4754fa60b88bb33991b288. [Reporting paper]() Recorded evaluation configuration: {"Model":"Gemini 2.5 Flash"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 10.3. Result ID: 6a4754fa60b88bb33991b279. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-5 Mini"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 10.3. Result ID: 6a4754fa60b88bb33991b28e. [Reporting paper]() Recorded evaluation configuration: {"Model":"Claude Sonnet 4.6"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 12.9. Result ID: 6a4754fa60b88bb33991b28f. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Claude Sonnet 4.6"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 13.9. Result ID: 6a4754fa60b88bb33991b27a. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"GPT-5 Mini"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 14.5. Result ID: 6a4754fa60b88bb33991b285. [Reporting paper]() Recorded evaluation configuration: {"Model":"Gemini 3.1 Flash-Lite"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 15.2. Result ID: 6a4754fa60b88bb33991b287. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Gemini 2.5 Flash"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 15.2. Result ID: 6a4754fa60b88bb33991b291. [Reporting paper]() Recorded evaluation configuration: {"Model":"Claude Haiku 4.5"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 15.5. Result ID: 6a4754fa60b88bb33991b27b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-4.1"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 16.8. Result ID: 6a4754fa60b88bb33991b284. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Gemini 3.1 Flash-Lite"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 19.4. Result ID: 6a4754fa60b88bb33991b28d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Claude Sonnet 4.6"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 26.5. Result ID: 6a4754fa60b88bb33991b292. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Claude Haiku 4.5"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 29. Result ID: 6a4754fa60b88bb33991b27e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-4.1 Mini"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 32.9. Result ID: 6a4754fa60b88bb33991b289. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Gemini 2.5 Flash"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 34.8. Result ID: 6a4754fa60b88bb33991b27c. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-4.1"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 37.4. Result ID: 6a4754fa60b88bb33991b27f. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-4.1 Mini"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 47.7. Result ID: 6a4754fa60b88bb33991b280. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"GPT-4.1 Mini"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 49. Result ID: 6a4754fa60b88bb33991b286. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Gemini 3.1 Flash-Lite"}. This record does not assert independent reproduction. ### ProvenanceGuard Error Rate (ER): 51. Result ID: 6a4754fa60b88bb33991b28a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Gemini 2.5 Flash-Lite"}. This record does not assert independent reproduction. ### ProvePrompt Error Rate (ER): 59.7. Result ID: 6a4754fa60b88bb33991b28b. [Reporting paper]() Recorded evaluation configuration: {"Model":"Gemini 2.5 Flash-Lite"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 64.8. Result ID: 6a4754fa60b88bb33991b27d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"GPT-4.1"}. This record does not assert independent reproduction. ### DirectPrompt Error Rate (ER): 66.8. Result ID: 6a4754fa60b88bb33991b28c. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Gemini 2.5 Flash-Lite"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.