# Zero-shot long-context reasoning on ZeroSCROLLS

Source: [Zero-shot long-context reasoning on ZeroSCROLLS](<https://www.sota2.com/research/sota/zero-shot-long-context-reasoning-on-zeroscrolls>)

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

Data updated: 2026-03-23T11:14:20.084Z

Evaluation of zero-shot reasoning capabilities on the ZeroSCROLLS dataset, focusing on throughput (Tokens, Speed) and latency 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 | Zero-shot long-context reasoning |
| Dataset (including any stated split) | ZeroSCROLLS |
| Metric | Average Score |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 18 of 18 |
| 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

- [Average Score](<https://www.sota2.com/research/sota/zero-shot-long-context-reasoning-on-zeroscrolls.md?metric=Average%20Score>): Higher is better; unit: SCALAR.

- [Tokens Processed](<https://www.sota2.com/research/sota/zero-shot-long-context-reasoning-on-zeroscrolls.md?metric=Tokens%20Processed>): Higher is better; unit: SCALAR.

- [Speedup Factor (1/τ)](<https://www.sota2.com/research/sota/zero-shot-long-context-reasoning-on-zeroscrolls.md?metric=Speedup%20Factor%20(1%2F%CF%84)>): Higher is better; unit: SCALAR.

- [Latency](<https://www.sota2.com/research/sota/zero-shot-long-context-reasoning-on-zeroscrolls.md?metric=Latency>): Lower is better; unit: TIME.

- [Speed Factor](<https://www.sota2.com/research/sota/zero-shot-long-context-reasoning-on-zeroscrolls.md?metric=Speed%20Factor>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Average Score | Result date | Report paper |
| --- | --- | --- | --- |
| LLMLingua-2 | 33.5 | Not recorded | 69c0c8d70636e5bf4afc394a |
| LLMLingua-2 | 33.4 | Not recorded | 69c0c8d70636e5bf4afc394a |
| LLMLingua-2-small | 33.4 | Not recorded | 69c0c8d70636e5bf4afc394a |
| LLMLingua-2-small | 33.3 | Not recorded | 69c0c8d70636e5bf4afc394a |
| LongLLMLingua | 33 | Not recorded | 69c0c8d70636e5bf4afc394a |
| LongLLMLingua | 32.7 | Not recorded | 69c0c8d70636e5bf4afc394a |
| Original Prompt | 32.5 | Not recorded | 69c0c8d70636e5bf4afc394a |
| BEAVER | 32.4 | Not recorded | 69c0c8d70636e5bf4afc394a |
| BEAVER | 32 | Not recorded | 69c0c8d70636e5bf4afc394a |
| LLMLingua | 30.7 | Not recorded | 69c0c8d70636e5bf4afc394a |
| LLMLingua | 27.2 | Not recorded | 69c0c8d70636e5bf4afc394a |
| SBERT | 24 | Not recorded | 69c0c8d70636e5bf4afc394a |
| OpenAI | 22.4 | Not recorded | 69c0c8d70636e5bf4afc394a |
| Selective-Context | 20.7 | Not recorded | 69c0c8d70636e5bf4afc394a |
| OpenAI | 20.6 | Not recorded | 69c0c8d70636e5bf4afc394a |
| SBERT | 20.5 | Not recorded | 69c0c8d70636e5bf4afc394a |
| Selective-Context | 19.4 | Not recorded | 69c0c8d70636e5bf4afc394a |
| Zero-Shot | 12.8 | Not recorded | 69c0c8d70636e5bf4afc394a |

### LLMLingua-2

Average Score: 33.5. Result ID: 69c1208247f34ce924a97e73.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

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

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

Recorded evaluation configuration: {"Token constraint":"3,000"}. This record does not assert independent reproduction.

### LLMLingua-2

Average Score: 33.4. Result ID: 69c1208247f34ce924a97e6b.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

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

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

Recorded evaluation configuration: {"Token constraint":"2,000"}. This record does not assert independent reproduction.

### LLMLingua-2-small

Average Score: 33.4. Result ID: 69c1208247f34ce924a97e72.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

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

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

Recorded evaluation configuration: {"Token constraint":"3,000"}. This record does not assert independent reproduction.

### LLMLingua-2-small

Average Score: 33.3. Result ID: 69c1208247f34ce924a97e6a.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

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

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

Recorded evaluation configuration: {"Token constraint":"2,000"}. This record does not assert independent reproduction.

### LongLLMLingua

Average Score: 33. Result ID: 69c1208247f34ce924a97e6e.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61c6d.md>)

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

Recorded evaluation configuration: {"Token constraint":"3,000"}. This record does not assert independent reproduction.

### LongLLMLingua

Average Score: 32.7. Result ID: 69c1208247f34ce924a97e66.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74701e209131cea61c6d.md>)

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

Recorded evaluation configuration: {"Token constraint":"2,000"}. This record does not assert independent reproduction.

### Original Prompt

Average Score: 32.5. Result ID: 69c1208247f34ce924a97e75.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

Recorded evaluation configuration: {"Configuration":"Reference Baseline"}. This record does not assert independent reproduction.

### BEAVER

Average Score: 32.4. Result ID: 69c1208247f34ce924a97e74.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

[Implementation](<https://github.com/JusperLee/BEAVER>)

Recorded evaluation configuration: {"Token constraint":"3,000"}. This record does not assert independent reproduction.

### BEAVER

Average Score: 32. Result ID: 69c1208247f34ce924a97e6c.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

[Implementation](<https://github.com/JusperLee/BEAVER>)

Recorded evaluation configuration: {"Token constraint":"2,000"}. This record does not assert independent reproduction.

### LLMLingua

Average Score: 30.7. Result ID: 69c1208247f34ce924a97e71.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

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

Recorded evaluation configuration: {"Token constraint":"3,000"}. This record does not assert independent reproduction.

### LLMLingua

Average Score: 27.2. Result ID: 69c1208247f34ce924a97e69.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

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

Recorded evaluation configuration: {"Token constraint":"2,000"}. This record does not assert independent reproduction.

### SBERT

Average Score: 24. Result ID: 69c1208247f34ce924a97e6f.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66441e209131cea0f5ab.md>)

[Implementation](<https://github.com/UKPLab/sentence-transformers>)

Recorded evaluation configuration: {"Token constraint":"3,000"}. This record does not assert independent reproduction.

### OpenAI

Average Score: 22.4. Result ID: 69c1208247f34ce924a97e70.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

Recorded evaluation configuration: {"Token constraint":"3,000"}. This record does not assert independent reproduction.

### Selective-Context

Average Score: 20.7. Result ID: 69c1208247f34ce924a97e6d.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

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

Recorded evaluation configuration: {"Token constraint":"3,000"}. This record does not assert independent reproduction.

### OpenAI

Average Score: 20.6. Result ID: 69c1208247f34ce924a97e68.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

Recorded evaluation configuration: {"Token constraint":"2,000"}. This record does not assert independent reproduction.

### SBERT

Average Score: 20.5. Result ID: 69c1208247f34ce924a97e67.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66441e209131cea0f5ab.md>)

[Implementation](<https://github.com/UKPLab/sentence-transformers>)

Recorded evaluation configuration: {"Token constraint":"2,000"}. This record does not assert independent reproduction.

### Selective-Context

Average Score: 19.4. Result ID: 69c1208247f34ce924a97e65.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

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

Recorded evaluation configuration: {"Token constraint":"2,000"}. This record does not assert independent reproduction.

### Zero-Shot

Average Score: 12.8. Result ID: 69c1208247f34ce924a97e76.

[Reporting paper](<https://www.sota2.com/research/paper/69c0c8d70636e5bf4afc394a.md>)

Recorded evaluation configuration: {"Configuration":"Reference Baseline"}. This record does not assert independent reproduction.

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