# Long-context Understanding on LongBench LLaMA-3.1-8B-Instruct (test)

Source: [Long-context Understanding on LongBench LLaMA-3.1-8B-Instruct (test)](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test>)

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

Data updated: 2026-04-10T00:18:13.797Z

Evaluation of the LLaMA-3.1-8B-Instruct model on the LongBench test set, reporting performance across various long-context understanding sub-tasks.

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 Understanding |
| Dataset (including any stated split) | LongBench LLaMA-3.1-8B-Instruct (test) |
| Metric | NrtvQA |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| 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

- [NrtvQA](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=NrtvQA>): Higher is better; unit: PERCENT.

- [Qasper](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=Qasper>): Higher is better; unit: PERCENT.

- [MF-en](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=MF-en>): Higher is better; unit: PERCENT.

- [HotpotQA](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=HotpotQA>): Higher is better; unit: PERCENT.

- [2Wiki](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=2Wiki>): Higher is better; unit: PERCENT.

- [Musique](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=Musique>): Higher is better; unit: PERCENT.

- [GovRep](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=GovRep>): Higher is better; unit: PERCENT.

- [QMSum](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=QMSum>): Higher is better; unit: PERCENT.

- [MNews](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=MNews>): Higher is better; unit: PERCENT.

- [TREC](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=TREC>): Higher is better; unit: PERCENT.

- [Trivia](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=Trivia>): Higher is better; unit: PERCENT.

- [SAMSum](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=SAMSum>): Higher is better; unit: PERCENT.

- [PCount](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=PCount>): Higher is better; unit: PERCENT.

- [PRe](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=PRe>): Higher is better; unit: PERCENT.

- [Lcc](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=Lcc>): Higher is better; unit: PERCENT.

- [RB-P](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=RB-P>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/long-context-understanding-on-longbench-llama-3-1-8b-instruct-test.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | NrtvQA | Result date | Report paper |
| --- | --- | --- | --- |
| StructKV | 32.8 | Not recorded | 69d732b57d0172140ec76cbb |
| StructKV | 32.45 | Not recorded | 69d732b57d0172140ec76cbb |
| SnapKV | 31.51 | Not recorded | 69d732b57d0172140ec76cbb |
| SnapKV | 31.15 | Not recorded | 69d732b57d0172140ec76cbb |
| FastKV | 30.54 | Not recorded | 69d732b57d0172140ec76cbb |
| FastKV | 30.26 | Not recorded | 69d732b57d0172140ec76cbb |
| Full-context | 30.21 | Not recorded | 69d732b57d0172140ec76cbb |
| StreamingLLM | 27.99 | Not recorded | 69d732b57d0172140ec76cbb |
| StreamingLLM | 26.05 | Not recorded | 69d732b57d0172140ec76cbb |
| GemFilter | 26.02 | Not recorded | 69d732b57d0172140ec76cbb |
| GemFilter | 24.36 | Not recorded | 69d732b57d0172140ec76cbb |
| PyramidInfer | 12.1 | Not recorded | 69d732b57d0172140ec76cbb |
| H2O | 9.8 | Not recorded | 69d732b57d0172140ec76cbb |
| H2O | 9.32 | Not recorded | 69d732b57d0172140ec76cbb |

### StructKV

NrtvQA: 32.8. Result ID: 69d8412347f34ce924c044af.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

Recorded evaluation configuration: {"Category":"Prefill-Aware","Prefill Budget":"60%","KV Budget":"20%"}. This record does not assert independent reproduction.

### StructKV

NrtvQA: 32.45. Result ID: 69d8412347f34ce924c044ae.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

Recorded evaluation configuration: {"Category":"Prefill-Aware","Prefill Budget":"60%","KV Budget":"10%"}. This record does not assert independent reproduction.

### SnapKV

NrtvQA: 31.51. Result ID: 69d8412347f34ce924c044a7.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73b21e209131cea5df89.md>)

[Implementation](<https://github.com/FasterDecoding/SnapKV>)

Recorded evaluation configuration: {"Category":"Decoding-Only","Prefill Budget":"100%","KV Budget":"10%"}. This record does not assert independent reproduction.

### SnapKV

NrtvQA: 31.15. Result ID: 69d8412347f34ce924c044a8.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73b21e209131cea5df89.md>)

[Implementation](<https://github.com/FasterDecoding/SnapKV>)

Recorded evaluation configuration: {"Category":"Decoding-Only","Prefill Budget":"100%","KV Budget":"20%"}. This record does not assert independent reproduction.

### FastKV

NrtvQA: 30.54. Result ID: 69d8412347f34ce924c044ac.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a41e209131cea79874.md>)

Recorded evaluation configuration: {"Category":"Prefill-Aware","Prefill Budget":"60%","KV Budget":"10%"}. This record does not assert independent reproduction.

### FastKV

NrtvQA: 30.26. Result ID: 69d8412347f34ce924c044ad.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa78a41e209131cea79874.md>)

Recorded evaluation configuration: {"Category":"Prefill-Aware","Prefill Budget":"60%","KV Budget":"20%"}. This record does not assert independent reproduction.

### Full-context

NrtvQA: 30.21. Result ID: 69d8412347f34ce924c044a2.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

Recorded evaluation configuration: {"Category":"Full-context","Prefill Budget":"100%","KV Budget":"100%"}. This record does not assert independent reproduction.

### StreamingLLM

NrtvQA: 27.99. Result ID: 69d8412347f34ce924c044a4.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d79.md>)

Recorded evaluation configuration: {"Category":"Decoding-Only","Prefill Budget":"100%","KV Budget":"20%"}. This record does not assert independent reproduction.

### StreamingLLM

NrtvQA: 26.05. Result ID: 69d8412347f34ce924c044a3.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72d71e209131cea58d79.md>)

Recorded evaluation configuration: {"Category":"Decoding-Only","Prefill Budget":"100%","KV Budget":"10%"}. This record does not assert independent reproduction.

### GemFilter

NrtvQA: 26.02. Result ID: 69d8412347f34ce924c044ab.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74fe1e209131cea651d5.md>)

Recorded evaluation configuration: {"Category":"Prefill-Aware","Prefill Budget":"70%","KV Budget":"20%"}. This record does not assert independent reproduction.

### GemFilter

NrtvQA: 24.36. Result ID: 69d8412347f34ce924c044aa.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74fe1e209131cea651d5.md>)

Recorded evaluation configuration: {"Category":"Prefill-Aware","Prefill Budget":"60%","KV Budget":"10%"}. This record does not assert independent reproduction.

### PyramidInfer

NrtvQA: 12.1. Result ID: 69d8412347f34ce924c044a9.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa73711e209131cea5c836.md>)

Recorded evaluation configuration: {"Category":"Prefill-Aware","Prefill Budget":"60%","KV Budget":"60%"}. This record does not assert independent reproduction.

### H2O

NrtvQA: 9.8. Result ID: 69d8412347f34ce924c044a6.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa718c1e209131cea51ead.md>)

[Implementation](<https://github.com/FMInference/H2O>)

Recorded evaluation configuration: {"Category":"Decoding-Only","Prefill Budget":"100%","KV Budget":"20%"}. This record does not assert independent reproduction.

### H2O

NrtvQA: 9.32. Result ID: 69d8412347f34ce924c044a5.

[Reporting paper](<https://www.sota2.com/research/paper/69d732b57d0172140ec76cbb.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa718c1e209131cea51ead.md>)

[Implementation](<https://github.com/FMInference/H2O>)

Recorded evaluation configuration: {"Category":"Decoding-Only","Prefill Budget":"100%","KV Budget":"10%"}. This record does not assert independent reproduction.

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