# Brain-to-Text Decoding on B2T '24 (test)

Source: [Brain-to-Text Decoding on B2T '24 (test)](<https://www.sota2.com/research/sota/brain-to-text-decoding-on-b2t-24-test>)

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

Data updated: 2026-03-17T09:05:04.333Z

Evaluation of Brain-to-Text Decoding performance on the test split of the B2T '24 dataset, reporting Word Error Rate (WER) alongside resource utilization (RAM, VRAM) and runtime efficiency (RTF).

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 | Brain-to-Text Decoding |
| Dataset (including any stated split) | B2T '24 (test) |
| Metric | WER |
| Unit | SCALAR |
| Direction | Lower is better |
| Results loaded | 3 of 3 |
| 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

- [WER](<https://www.sota2.com/research/sota/brain-to-text-decoding-on-b2t-24-test.md?metric=WER>): Lower is better; unit: SCALAR.

- [RAM Usage](<https://www.sota2.com/research/sota/brain-to-text-decoding-on-b2t-24-test.md?metric=RAM%20Usage>): Lower is better; unit: SIZE.

- [VRAM Usage](<https://www.sota2.com/research/sota/brain-to-text-decoding-on-b2t-24-test.md?metric=VRAM%20Usage>): Lower is better; unit: SIZE.

- [RTF](<https://www.sota2.com/research/sota/brain-to-text-decoding-on-b2t-24-test.md?metric=RTF>): Lower is better; unit: SCALAR.

## Recorded results

| Method | WER | Result date | Report paper |
| --- | --- | --- | --- |
| LightBeam | 9.37 | Not recorded | 69b8dff80636e5bf4afbe93c |
| WFST-Based (Re-impl.) | 9.71 | Not recorded | 69b8dff80636e5bf4afbe93c |
| WFST-Based (Orig.) | 9.76 | Not recorded | 69b8dff80636e5bf4afbe93c |

### LightBeam

WER: 9.37. Result ID: 69b9193147f34ce924a1dcb1.

[Reporting paper](<https://www.sota2.com/research/paper/69b8dff80636e5bf4afbe93c.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69b8dff80636e5bf4afbe93c.md>)

Recorded evaluation configuration: {"Encoder":"GRU","Hardware":"GeForce RTX 5090 GPU"}. This record does not assert independent reproduction.

### WFST-Based (Re-impl.)

WER: 9.71. Result ID: 69b9193147f34ce924a1dcb0.

[Reporting paper](<https://www.sota2.com/research/paper/69b8dff80636e5bf4afbe93c.md>)

Recorded evaluation configuration: {"Encoder":"GRU","Hardware":"GeForce RTX 5090 GPU"}. This record does not assert independent reproduction.

### WFST-Based (Orig.)

WER: 9.76. Result ID: 69b9193147f34ce924a1dcaf.

[Reporting paper](<https://www.sota2.com/research/paper/69b8dff80636e5bf4afbe93c.md>)

Recorded evaluation configuration: {"Encoder":"GRU"}. This record does not assert independent reproduction.

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