# Synthetic Tasks on LongBench Synthetic Tasks

Source: [Synthetic Tasks on LongBench Synthetic Tasks](<https://www.sota2.com/research/sota/synthetic-tasks-on-longbench-synthetic-tasks>)

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

Data updated: 2026-05-19T08:10:26.032Z

Evaluation of language models on the Synthetic Tasks subset of the LongBench benchmark, measured using Accuracy.

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 | Synthetic Tasks |
| Dataset (including any stated split) | LongBench Synthetic Tasks |
| Metric | Accuracy |
| 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

- [Accuracy](<https://www.sota2.com/research/sota/synthetic-tasks-on-longbench-synthetic-tasks.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| baseline | 67.78 | Not recorded | 68fa77a11e209131cea747e1 |
| KiVi | 67.74 | Not recorded | 68fa77a11e209131cea747e1 |
| LogQuant | 67.68 | Not recorded | 68fa77a11e209131cea747e1 |
| LogQuant | 52.51 | Not recorded | 68fa77a11e209131cea747e1 |
| KiVi | 45.02 | Not recorded | 68fa77a11e209131cea747e1 |

### baseline

Accuracy: 67.78. Result ID: 6a0c1af0fbc373592c020f94.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a11e209131cea747e1.md>)

Recorded evaluation configuration: {"Precision":"16-bit","Backbone":"Llama3.1-8B"}. This record does not assert independent reproduction.

### KiVi

Accuracy: 67.74. Result ID: 6a0c1af0fbc373592c020f91.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a11e209131cea747e1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa744b1e209131cea60dab.md>)

Recorded evaluation configuration: {"Precision":"4-bit","Backbone":"Llama3.1-8B"}. This record does not assert independent reproduction.

### LogQuant

Accuracy: 67.68. Result ID: 6a0c1af0fbc373592c020f93.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a11e209131cea747e1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77a11e209131cea747e1.md>)

Recorded evaluation configuration: {"Precision":"4-bit","Backbone":"Llama3.1-8B"}. This record does not assert independent reproduction.

### LogQuant

Accuracy: 52.51. Result ID: 6a0c1af0fbc373592c020f92.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a11e209131cea747e1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77a11e209131cea747e1.md>)

Recorded evaluation configuration: {"Precision":"2-bit","Backbone":"Llama3.1-8B"}. This record does not assert independent reproduction.

### KiVi

Accuracy: 45.02. Result ID: 6a0c1af0fbc373592c020f90.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77a11e209131cea747e1.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa744b1e209131cea60dab.md>)

Recorded evaluation configuration: {"Precision":"2-bit","Backbone":"Llama3.1-8B"}. This record does not assert independent reproduction.

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