# Next Token Prediction on Observed Antibody Sequences (OAS) processed subset

Source: [Next Token Prediction on Observed Antibody Sequences (OAS) processed subset](<https://www.sota2.com/research/sota/next-token-prediction-on-observed-antibody-sequences-oas-processed-subset>)

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

Data updated: 2026-02-26T05:37:50.725Z

Evaluation of sequence models performing Next Token Prediction on the processed subset of the Observed Antibody Sequences (OAS) dataset, assessed using metrics such as Bits Per Dimension, Top-1/Top-10 Error, and Validation Loss.

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 | Next Token Prediction |
| Dataset (including any stated split) | Observed Antibody Sequences (OAS) processed subset |
| Metric | Bits Per Dimension |
| Unit | SCALAR |
| Direction | Lower 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

- [Bits Per Dimension](<https://www.sota2.com/research/sota/next-token-prediction-on-observed-antibody-sequences-oas-processed-subset.md?metric=Bits%20Per%20Dimension>): Lower is better; unit: SCALAR.

- [Top-1 Error](<https://www.sota2.com/research/sota/next-token-prediction-on-observed-antibody-sequences-oas-processed-subset.md?metric=Top-1%20Error>): Lower is better; unit: PERCENT.

- [Top-10 Error](<https://www.sota2.com/research/sota/next-token-prediction-on-observed-antibody-sequences-oas-processed-subset.md?metric=Top-10%20Error>): Lower is better; unit: PERCENT.

- [Validation Loss](<https://www.sota2.com/research/sota/next-token-prediction-on-observed-antibody-sequences-oas-processed-subset.md?metric=Validation%20Loss>): Lower is better; unit: SCALAR.

## Recorded results

| Method | Bits Per Dimension | Result date | Report paper |
| --- | --- | --- | --- |
| Mistral 94M | 1.62 | Not recorded | 68fa74401e209131cea6086f |
| Mistral 212M | 2.06 | Not recorded | 68fa74401e209131cea6086f |
| Mistral 377M | 2.41 | Not recorded | 68fa74401e209131cea6086f |
| Random Guess | 4.86 | Not recorded | 68fa74401e209131cea6086f |

### Mistral 94M

Bits Per Dimension: 1.62. Result ID: 699336ef3a6c3148491479e6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74401e209131cea6086f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74401e209131cea6086f.md>)

Recorded evaluation configuration: {"Architecture":"Mistral 7B (scaled-down)","Parameters":"94M","Quantization":"Post-training quantization"}. This record does not assert independent reproduction.

### Mistral 212M

Bits Per Dimension: 2.06. Result ID: 699336ef3a6c3148491479e5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74401e209131cea6086f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74401e209131cea6086f.md>)

Recorded evaluation configuration: {"Architecture":"Mistral 7B (scaled-down)","Parameters":"212M","Quantization":"Post-training quantization"}. This record does not assert independent reproduction.

### Mistral 377M

Bits Per Dimension: 2.41. Result ID: 699336ef3a6c3148491479e4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74401e209131cea6086f.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa74401e209131cea6086f.md>)

Recorded evaluation configuration: {"Architecture":"Mistral 7B (scaled-down)","Parameters":"377M","Quantization":"Post-training quantization"}. This record does not assert independent reproduction.

### Random Guess

Bits Per Dimension: 4.86. Result ID: 699336ef3a6c3148491479e7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa74401e209131cea6086f.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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