# Long-context Information Extraction on RULER 4K-32K Average Source: [Long-context Information Extraction on RULER 4K-32K Average]() Collection index: [llms.txt]() Data updated: 2026-05-13T09:34:13.694Z Evaluation of model performance on the RULER dataset covering context lengths from 4,000 to 32,000 tokens, reporting CWE and FWE scores. 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 | Long-context Information Extraction | | Dataset (including any stated split) | RULER 4K-32K Average | | Metric | CWE Score | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 6 of 6 | | 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 - [CWE Score](): Higher is better; unit: PERCENT. - [FWE Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | CWE Score | Result date | Report paper | | --- | --- | --- | --- | | Training–Inference Consistent Segmented Execution | 46.39 | Not recorded | 6a0405c9dbc836b87d8277fc | | LLaMA2-7B-32K (Vanilla Self-Attention) | 32.94 | Not recorded | 6a0405c9dbc836b87d8277fc | | MInference | 32.94 | Not recorded | 6a0405c9dbc836b87d8277fc | | DuoAttention | 32.33 | Not recorded | 6a0405c9dbc836b87d8277fc | | StreamingLLM | 27.78 | Not recorded | 6a0405c9dbc836b87d8277fc | | CCA-attention | 24.9 | Not recorded | 6a0405c9dbc836b87d8277fc | ### Training–Inference Consistent Segmented Execution CWE Score: 46.39. Result ID: 6a0444c747f34ce924ea7307. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaMA2-7B-32K"}. This record does not assert independent reproduction. ### LLaMA2-7B-32K (Vanilla Self-Attention) CWE Score: 32.94. Result ID: 6a0444c747f34ce924ea7302. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaMA2-7B-32K","Attention":"Vanilla Self-Attention"}. This record does not assert independent reproduction. ### MInference CWE Score: 32.94. Result ID: 6a0444c747f34ce924ea7304. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"LLaMA2-7B-32K"}. This record does not assert independent reproduction. ### DuoAttention CWE Score: 32.33. Result ID: 6a0444c747f34ce924ea7306. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaMA2-7B-32K"}. This record does not assert independent reproduction. ### StreamingLLM CWE Score: 27.78. Result ID: 6a0444c747f34ce924ea7303. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaMA2-7B-32K"}. This record does not assert independent reproduction. ### CCA-attention CWE Score: 24.9. Result ID: 6a0444c747f34ce924ea7305. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"LLaMA2-7B-32K"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.