# Answer-level hallucination detection on HalluRAG Source: [Answer-level hallucination detection on HalluRAG]() Collection index: [llms.txt]() Data updated: 2026-07-01T07:11:30.442Z Evaluation of model performance on detecting hallucinations specifically at the answer level using the HalluRAG 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 | Answer-level hallucination detection | | Dataset (including any stated split) | HalluRAG | | Metric | AP | | 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 - [AP](): Higher is better; unit: PERCENT. - [AUROC](): Higher is better; unit: PERCENT. ## Recorded results | Method | AP | Result date | Report paper | | --- | --- | --- | --- | | CORTEX | 80.1 | Not recorded | 6a449f02e0224bd4f4992fc1 | | RAGLens | 80.06 | Not recorded | 6a449f02e0224bd4f4992fc1 | | RAGLens | 79.94 | Not recorded | 6a449f02e0224bd4f4992fc1 | | SAPLMA | 79.22 | Not recorded | 6a449f02e0224bd4f4992fc1 | | CORTEX | 79.07 | Not recorded | 6a449f02e0224bd4f4992fc1 | | SAPLMA | 78.67 | Not recorded | 6a449f02e0224bd4f4992fc1 | | CORTEX | 77.95 | Not recorded | 6a449f02e0224bd4f4992fc1 | | RAGLens | 76.69 | Not recorded | 6a449f02e0224bd4f4992fc1 | | SAPLMA | 75.29 | Not recorded | 6a449f02e0224bd4f4992fc1 | | ICR Probe | 40.77 | Not recorded | 6a449f02e0224bd4f4992fc1 | | LLM-Check | 40.59 | Not recorded | 6a449f02e0224bd4f4992fc1 | | LLM-Check | 39.56 | Not recorded | 6a449f02e0224bd4f4992fc1 | | ICR Probe | 32.86 | Not recorded | 6a449f02e0224bd4f4992fc1 | | NLL | 31.84 | Not recorded | 6a449f02e0224bd4f4992fc1 | | LLM-Check | 30.52 | Not recorded | 6a449f02e0224bd4f4992fc1 | | NLL | 29.69 | Not recorded | 6a449f02e0224bd4f4992fc1 | | ICR Probe | 21.57 | Not recorded | 6a449f02e0224bd4f4992fc1 | | NLL | 20.55 | Not recorded | 6a449f02e0224bd4f4992fc1 | ### CORTEX AP: 80.1. Result ID: 6a44bda173d097d845be1b60. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Mistral"}. This record does not assert independent reproduction. ### RAGLens AP: 80.06. Result ID: 6a44bda173d097d845be1b5f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Mistral"}. This record does not assert independent reproduction. ### RAGLens AP: 79.94. Result ID: 6a44bda173d097d845be1b53. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Llama"}. This record does not assert independent reproduction. ### SAPLMA AP: 79.22. Result ID: 6a44bda173d097d845be1b52. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Llama"}. This record does not assert independent reproduction. ### CORTEX AP: 79.07. Result ID: 6a44bda173d097d845be1b5a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen"}. This record does not assert independent reproduction. ### SAPLMA AP: 78.67. Result ID: 6a44bda173d097d845be1b58. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen"}. This record does not assert independent reproduction. ### CORTEX AP: 77.95. Result ID: 6a44bda173d097d845be1b54. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Llama"}. This record does not assert independent reproduction. ### RAGLens AP: 76.69. Result ID: 6a44bda173d097d845be1b59. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen"}. This record does not assert independent reproduction. ### SAPLMA AP: 75.29. Result ID: 6a44bda173d097d845be1b5e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Mistral"}. This record does not assert independent reproduction. ### ICR Probe AP: 40.77. Result ID: 6a44bda173d097d845be1b51. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Llama"}. This record does not assert independent reproduction. ### LLM-Check AP: 40.59. Result ID: 6a44bda173d097d845be1b5c. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"Mistral"}. This record does not assert independent reproduction. ### LLM-Check AP: 39.56. Result ID: 6a44bda173d097d845be1b56. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen"}. This record does not assert independent reproduction. ### ICR Probe AP: 32.86. Result ID: 6a44bda173d097d845be1b57. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen"}. This record does not assert independent reproduction. ### NLL AP: 31.84. Result ID: 6a44bda173d097d845be1b55. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Qwen"}. This record does not assert independent reproduction. ### LLM-Check AP: 30.52. Result ID: 6a44bda173d097d845be1b50. [Reporting paper]() Recorded evaluation configuration: {"Backbone LLM":"Llama"}. This record does not assert independent reproduction. ### NLL AP: 29.69. Result ID: 6a44bda173d097d845be1b5b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Mistral"}. This record does not assert independent reproduction. ### ICR Probe AP: 21.57. Result ID: 6a44bda173d097d845be1b5d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Mistral"}. This record does not assert independent reproduction. ### NLL AP: 20.55. Result ID: 6a44bda173d097d845be1b4f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone LLM":"Llama"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.