# Question Answering on SQuAD (CA, ASR) Source: [Question Answering on SQuAD (CA, ASR)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:34:11.590Z Evaluation of model performance on the Stanford Question Answering Dataset (SQuAD) using CA (Correct Answer?) and ASR (Answer Success Rate) metrics. 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 | Question Answering | | Dataset (including any stated split) | SQuAD | | Metric | Accuracy (CA) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 8 of 8 | | 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 (CA)](): Higher is better; unit: PERCENT. - [Answer Success Rate (ASR)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (CA) | Result date | Report paper | | --- | --- | --- | --- | | Control Model | 78.5 | Not recorded | 6976f4fda2b28e761d87c1cd | | Control Model | 78.4 | Not recorded | 6976f4fda2b28e761d87c1cd | | Control Model | 78.2 | Not recorded | 6976f4fda2b28e761d87c1cd | | Control Model | 78.2 | Not recorded | 6976f4fda2b28e761d87c1cd | | LRAM | 76.3 | Not recorded | 6976f4fda2b28e761d87c1cd | | LRAM | 76.3 | Not recorded | 6976f4fda2b28e761d87c1cd | | Compression-Activated Model | 76.3 | Not recorded | 6976f4fda2b28e761d87c1cd | | Compression-Activated Model | 76.3 | Not recorded | 6976f4fda2b28e761d87c1cd | ### Control Model Accuracy (CA): 78.5. Result ID: 69991ec55698665ec485a942. [Reporting paper]() Recorded evaluation configuration: {"Rank":728,"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### Control Model Accuracy (CA): 78.4. Result ID: 69991ec55698665ec485a944. [Reporting paper]() Recorded evaluation configuration: {"Rank":500,"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### Control Model Accuracy (CA): 78.2. Result ID: 69991ec55698665ec485a946. [Reporting paper]() Recorded evaluation configuration: {"Pruned %":"0%","Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### Control Model Accuracy (CA): 78.2. Result ID: 69991ec55698665ec485a948. [Reporting paper]() Recorded evaluation configuration: {"Pruned %":"5%","Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### LRAM Accuracy (CA): 76.3. Result ID: 69991ec55698665ec485a943. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Rank":728,"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### LRAM Accuracy (CA): 76.3. Result ID: 69991ec55698665ec485a945. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Rank":500,"Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### Compression-Activated Model Accuracy (CA): 76.3. Result ID: 69991ec55698665ec485a947. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Pruned %":"0%","Backbone":"RoBERTa"}. This record does not assert independent reproduction. ### Compression-Activated Model Accuracy (CA): 76.3. Result ID: 69991ec55698665ec485a949. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Pruned %":"5%","Backbone":"RoBERTa"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.