# Needle-In-A-Haystack Retrieval on NIAH Single 2

Source: [Needle-In-A-Haystack Retrieval on NIAH Single 2](<https://www.sota2.com/research/sota/needle-in-a-haystack-retrieval-on-niah-single-2>)

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

Data updated: 2026-04-08T09:10:14.660Z

Evaluation of Needle-In-A-Haystack Retrieval performance on the NIAH Single 2 dataset, measuring success rates across context window sizes of 4K, 8K, 16K, and 32K tokens.

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 | Needle-In-A-Haystack Retrieval |
| Dataset (including any stated split) | NIAH Single 2 |
| Metric | Success Rate (4K Context) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 2 of 2 |
| 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

- [Success Rate (4K Context)](<https://www.sota2.com/research/sota/needle-in-a-haystack-retrieval-on-niah-single-2.md?metric=Success%20Rate%20(4K%20Context)>): Higher is better; unit: PERCENT.

- [Success Rate (8K Context)](<https://www.sota2.com/research/sota/needle-in-a-haystack-retrieval-on-niah-single-2.md?metric=Success%20Rate%20(8K%20Context)>): Higher is better; unit: PERCENT.

- [Success Rate (16K Context)](<https://www.sota2.com/research/sota/needle-in-a-haystack-retrieval-on-niah-single-2.md?metric=Success%20Rate%20(16K%20Context)>): Higher is better; unit: PERCENT.

- [Success Rate (32K Context)](<https://www.sota2.com/research/sota/needle-in-a-haystack-retrieval-on-niah-single-2.md?metric=Success%20Rate%20(32K%20Context)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Success Rate (4K Context) | Result date | Report paper |
| --- | --- | --- | --- |
| StateX | 94 | Not recorded | 68fa7bac1e209131cea89ce8 |
| Vanilla | 83 | Not recorded | 68fa7bac1e209131cea89ce8 |

### StateX

Success Rate (4K Context): 94. Result ID: 69d61b6647f34ce924bee34c.

[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":"Mamba2"}. This record does not assert independent reproduction.

### Vanilla

Success Rate (4K Context): 83. Result ID: 69d61b6647f34ce924bee34b.

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

Recorded evaluation configuration: {"Architecture":"Mamba2"}. This record does not assert independent reproduction.

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