# Recall-intensive Question Answering on SWDE, SQuAD, TQA, NQ, Drop (2K context, test)

Source: [Recall-intensive Question Answering on SWDE, SQuAD, TQA, NQ, Drop (2K context, test)](<https://www.sota2.com/research/sota/recall-intensive-question-answering-on-swde-squad-tqa-nq-drop-2k-context-test>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-04-08T09:10:14.683Z

Evaluation of Recall-intensive Question Answering performance across a suite of datasets (SWDE, SQuAD, TQA, NQ, Drop), where all inputs are truncated to a maximum context length of 2048 tokens, using the test splits.

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 | Recall-intensive Question Answering |
| Dataset (including any stated split) | Recall-intensive tasks (SWDE, SQuAD, TQA, NQ, Drop) truncated to 2K tokens (test) |
| Metric | SWDE |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [SWDE](<https://www.sota2.com/research/sota/recall-intensive-question-answering-on-swde-squad-tqa-nq-drop-2k-context-test.md?metric=SWDE>): Higher is better; unit: PERCENT.

- [SQuAD](<https://www.sota2.com/research/sota/recall-intensive-question-answering-on-swde-squad-tqa-nq-drop-2k-context-test.md?metric=SQuAD>): Higher is better; unit: PERCENT.

- [TQA](<https://www.sota2.com/research/sota/recall-intensive-question-answering-on-swde-squad-tqa-nq-drop-2k-context-test.md?metric=TQA>): Higher is better; unit: PERCENT.

- [NQ](<https://www.sota2.com/research/sota/recall-intensive-question-answering-on-swde-squad-tqa-nq-drop-2k-context-test.md?metric=NQ>): Higher is better; unit: PERCENT.

- [Drop](<https://www.sota2.com/research/sota/recall-intensive-question-answering-on-swde-squad-tqa-nq-drop-2k-context-test.md?metric=Drop>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/recall-intensive-question-answering-on-swde-squad-tqa-nq-drop-2k-context-test.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | SWDE | Result date | Report paper |
| --- | --- | --- | --- |
| StateX | 56.1 | Not recorded | 68fa7bac1e209131cea89ce8 |
| Vanilla | 54.1 | Not recorded | 68fa7bac1e209131cea89ce8 |
| StateX | 50.3 | Not recorded | 68fa7bac1e209131cea89ce8 |
| Vanilla | 47.1 | Not recorded | 68fa7bac1e209131cea89ce8 |
| MoM | 34.4 | Not recorded | 68fa7bac1e209131cea89ce8 |

### StateX

SWDE: 56.1. Result ID: 69d61b6647f34ce924bee350.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bac1e209131cea89ce8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bac1e209131cea89ce8.md>)

Recorded evaluation configuration: {"Architecture Family":"State Space Model — Mamba2","Params":"1.350B","Total State":"37.44M"}. This record does not assert independent reproduction.

### Vanilla

SWDE: 54.1. Result ID: 69d61b6647f34ce924bee34f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bac1e209131cea89ce8.md>)

Recorded evaluation configuration: {"Architecture Family":"State Space Model — Mamba2","Params":"1.343B","Total State":"24.96M"}. This record does not assert independent reproduction.

### StateX

SWDE: 50.3. Result ID: 69d61b6647f34ce924bee34e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bac1e209131cea89ce8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7bac1e209131cea89ce8.md>)

Recorded evaluation configuration: {"Architecture Family":"Linear Attention — GLA","Params":"1.365B","Total State":"18.87M"}. This record does not assert independent reproduction.

### Vanilla

SWDE: 47.1. Result ID: 69d61b6647f34ce924bee34d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bac1e209131cea89ce8.md>)

Recorded evaluation configuration: {"Architecture Family":"Linear Attention — GLA","Params":"1.365B","Total State":"12.58M"}. This record does not assert independent reproduction.

### MoM

SWDE: 34.4. Result ID: 69d61b6647f34ce924bee351.

[Reporting paper](<https://www.sota2.com/research/paper/68fa7bac1e209131cea89ce8.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa7c001e209131cea8bcea.md>)

Recorded evaluation configuration: {"Architecture Family":"Sparse Model – MoM","Params":"1.552B","Total State":"31.45M"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.