# Foundation Model Selection on 100 diverse natural language user queries (test)

Source: [Foundation Model Selection on 100 diverse natural language user queries (test)](<https://www.sota2.com/research/sota/foundation-model-selection-on-100-diverse-natural-language-user-queries-test>)

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

Data updated: 2026-03-12T08:25:08.258Z

Evaluation of model selection performance using 100 diverse natural language user queries (test split), measured by ranking metrics such as Avg Top-1, Top-1 Hit, and MRR.

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 | Foundation Model Selection |
| Dataset (including any stated split) | 100 diverse natural language user queries (test) |
| Metric | Average Top-1 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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 Top-1](<https://www.sota2.com/research/sota/foundation-model-selection-on-100-diverse-natural-language-user-queries-test.md?metric=Average%20Top-1>): Higher is better; unit: PERCENT.

- [Average Set Score](<https://www.sota2.com/research/sota/foundation-model-selection-on-100-diverse-natural-language-user-queries-test.md?metric=Average%20Set%20Score>): Higher is better; unit: PERCENT.

- [Top-1 Hit Rate](<https://www.sota2.com/research/sota/foundation-model-selection-on-100-diverse-natural-language-user-queries-test.md?metric=Top-1%20Hit%20Rate>): Higher is better; unit: PERCENT.

- [HQ Hit Rate](<https://www.sota2.com/research/sota/foundation-model-selection-on-100-diverse-natural-language-user-queries-test.md?metric=HQ%20Hit%20Rate>): Higher is better; unit: PERCENT.

- [Mean Reciprocal Rank (MRR)](<https://www.sota2.com/research/sota/foundation-model-selection-on-100-diverse-natural-language-user-queries-test.md?metric=Mean%20Reciprocal%20Rank%20(MRR)>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Average Top-1 | Result date | Report paper |
| --- | --- | --- | --- |
| REMSA | 75.76 | Not recorded | 6923e67dd6b0dd334292ef85 |
| REMSA-Naive | 72.67 | Not recorded | 6923e67dd6b0dd334292ef85 |
| UNSTR.-RAG | 71.23 | Not recorded | 6923e67dd6b0dd334292ef85 |
| DB-Retrieval | 67.37 | Not recorded | 6923e67dd6b0dd334292ef85 |

### REMSA

Average Top-1: 75.76. Result ID: 69b2784b6dbb7075b95d061a.

[Reporting paper](<https://www.sota2.com/research/paper/6923e67dd6b0dd334292ef85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6923e67dd6b0dd334292ef85.md>)

Recorded evaluation configuration: {"LLM Backbone":"GPT-4.1"}. This record does not assert independent reproduction.

### REMSA-Naive

Average Top-1: 72.67. Result ID: 69b2784b6dbb7075b95d061b.

[Reporting paper](<https://www.sota2.com/research/paper/6923e67dd6b0dd334292ef85.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6923e67dd6b0dd334292ef85.md>)

Recorded evaluation configuration: {"LLM Backbone":"GPT-4.1"}. This record does not assert independent reproduction.

### UNSTR.-RAG

Average Top-1: 71.23. Result ID: 69b2784b6dbb7075b95d061d.

[Reporting paper](<https://www.sota2.com/research/paper/6923e67dd6b0dd334292ef85.md>)

Recorded evaluation configuration: {"LLM Backbone":"GPT-4.1"}. This record does not assert independent reproduction.

### DB-Retrieval

Average Top-1: 67.37. Result ID: 69b2784b6dbb7075b95d061c.

[Reporting paper](<https://www.sota2.com/research/paper/6923e67dd6b0dd334292ef85.md>)

Recorded evaluation configuration: {"LLM Backbone":"GPT-4.1"}. This record does not assert independent reproduction.

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