# Question Answering on LongBench-SingleDoc (test)

Source: [Question Answering on LongBench-SingleDoc (test)](<https://www.sota2.com/research/sota/question-answering-on-longbench-singledoc-test>)

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

Data updated: 2026-02-26T06:04:04.394Z

Evaluation of Question Answering performance on the Single Document subset of the LongBench benchmark (test split), focusing on metrics related to context consumption.

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) | LongBench-SingleDoc (test) |
| Metric | Score (2k tokens) |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 9 of 9 |
| 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

- [Score (2k tokens)](<https://www.sota2.com/research/sota/question-answering-on-longbench-singledoc-test.md?metric=Score%20(2k%20tokens)>): Higher is better; unit: SCALAR.

- [Score (3k tokens)](<https://www.sota2.com/research/sota/question-answering-on-longbench-singledoc-test.md?metric=Score%20(3k%20tokens)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Score (2k tokens) | Result date | Report paper |
| --- | --- | --- | --- |
| LLMLingua-2 | 26.8 | Not recorded | 68fa74701e209131cea61cb2 |
| LLMLingua-2-small | 25.3 | Not recorded | 68fa74701e209131cea61cb2 |
| Original Prompt | 24.5 | Not recorded | 68fa74701e209131cea61cb2 |
| Selective-Context | 22 | Not recorded | 68fa74701e209131cea61cb2 |
| LLMLingua | 19.5 | Not recorded | 68fa74701e209131cea61cb2 |

### LLMLingua-2

Score (2k tokens): 26.8. Result ID: 699348bd3a6c31484914e9d1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61cb2.md>)

[Implementation](<https://aka.ms/LLMLingua-2>)

Recorded evaluation configuration: {"Target LLM":"Mistral-7B","Constraint Level":"2,000-token","Tokens":"1,967","Speedup (1/τ)":"7.4x"}. This record does not assert independent reproduction.

### LLMLingua-2-small

Score (2k tokens): 25.3. Result ID: 699348bd3a6c31484914e9cf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61cb2.md>)

[Implementation](<https://aka.ms/LLMLingua-2>)

Recorded evaluation configuration: {"Target LLM":"Mistral-7B","Constraint Level":"2,000-token","Tokens":"1,949","Speedup (1/τ)":"7.4x"}. This record does not assert independent reproduction.

### Original Prompt

Score (2k tokens): 24.5. Result ID: 699348bd3a6c31484914e9d3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61cb2.md>)

Recorded evaluation configuration: {"Target LLM":"Mistral-7B","Tokens":"14,511"}. This record does not assert independent reproduction.

### Selective-Context

Score (2k tokens): 22. Result ID: 699348bd3a6c31484914e9cb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d81e209131cea4d9ef.md>)

Recorded evaluation configuration: {"Target LLM":"Mistral-7B","Constraint Level":"2,000-token","Tokens":"2,038","Speedup (1/τ)":"7.1x"}. This record does not assert independent reproduction.

### LLMLingua

Score (2k tokens): 19.5. Result ID: 699348bd3a6c31484914e9cd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71621e209131cea5103a.md>)

Recorded evaluation configuration: {"Target LLM":"Mistral-7B","Constraint Level":"2,000-token","Tokens":"2,054","Speedup (1/τ)":"7.1x"}. This record does not assert independent reproduction.

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