# Abstention and Hallucination Evaluation on Controlled 50-item Evaluation Set (R1-R5) Source: [Abstention and Hallucination Evaluation on Controlled 50-item Evaluation Set (R1-R5)]() Collection index: [llms.txt]() Data updated: 2026-04-09T23:07:55.742Z Evaluation of model performance across five defined epistemic regimes (R1–R5), specifically measuring Abstention rates, Hallucination rates, and overall 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 | Abstention and Hallucination Evaluation | | Dataset (including any stated split) | Controlled 50-item Evaluation set epistemic regimes R1–R5 | | Metric | Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 12 of 12 | | 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 - [Accuracy](): Higher is better; unit: PERCENT. - [Hallucination Rate](): Lower is better; unit: PERCENT. - [Abstention Rate](): Lower is better; unit: PERCENT. ## Recorded results | Method | Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Instruction | 98 | Not recorded | 69d732617d0172140ec76957 | | Composite Abstention Architecture | 98 | Not recorded | 69d732617d0172140ec76957 | | Instruction | 98 | Not recorded | 69d732617d0172140ec76957 | | Composite Abstention Architecture | 98 | Not recorded | 69d732617d0172140ec76957 | | Composite Abstention Architecture | 96 | Not recorded | 69d732617d0172140ec76957 | | Instruction | 94 | Not recorded | 69d732617d0172140ec76957 | | Gate | 82 | Not recorded | 69d732617d0172140ec76957 | | Gate | 80 | Not recorded | 69d732617d0172140ec76957 | | Gate | 76 | Not recorded | 69d732617d0172140ec76957 | | Baseline | 70 | Not recorded | 69d732617d0172140ec76957 | | Baseline | 62 | Not recorded | 69d732617d0172140ec76957 | | Baseline | 50 | Not recorded | 69d732617d0172140ec76957 | ### Instruction Accuracy: 98. Result ID: 69d8312a47f34ce924bfd085. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-4o-mini"}. This record does not assert independent reproduction. ### Composite Abstention Architecture Accuracy: 98. Result ID: 69d8312a47f34ce924bfd087. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-4o-mini"}. This record does not assert independent reproduction. ### Instruction Accuracy: 98. Result ID: 69d8312a47f34ce924bfd089. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-4o"}. This record does not assert independent reproduction. ### Composite Abstention Architecture Accuracy: 98. Result ID: 69d8312a47f34ce924bfd08b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-4o"}. This record does not assert independent reproduction. ### Composite Abstention Architecture Accuracy: 96. Result ID: 69d8312a47f34ce924bfd08f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-3.5-turbo"}. This record does not assert independent reproduction. ### Instruction Accuracy: 94. Result ID: 69d8312a47f34ce924bfd08d. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-3.5-turbo"}. This record does not assert independent reproduction. ### Gate Accuracy: 82. Result ID: 69d8312a47f34ce924bfd08a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-4o"}. This record does not assert independent reproduction. ### Gate Accuracy: 80. Result ID: 69d8312a47f34ce924bfd086. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-4o-mini"}. This record does not assert independent reproduction. ### Gate Accuracy: 76. Result ID: 69d8312a47f34ce924bfd08e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"GPT-3.5-turbo"}. This record does not assert independent reproduction. ### Baseline Accuracy: 70. Result ID: 69d8312a47f34ce924bfd088. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-4o"}. This record does not assert independent reproduction. ### Baseline Accuracy: 62. Result ID: 69d8312a47f34ce924bfd084. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-4o-mini"}. This record does not assert independent reproduction. ### Baseline Accuracy: 50. Result ID: 69d8312a47f34ce924bfd08c. [Reporting paper]() Recorded evaluation configuration: {"Model":"GPT-3.5-turbo"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.