# Sentiment Classification on Stanford Sentiment Treebank SST-5 (test)

Source: [Sentiment Classification on Stanford Sentiment Treebank SST-5 (test)](<https://www.sota2.com/research/sota/sentiment-classification-on-stanford-sentiment-treebank-sst-5-test>)

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

Data updated: 2026-04-21T08:33:52.414Z

Evaluation of sentiment classification performance on the test split of the Stanford Sentiment Treebank (SST-5), measured by 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 | Sentiment Classification |
| Dataset (including any stated split) | Stanford Sentiment Treebank SST-5 (test) |
| Metric | Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 16 of 16 |
| 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/sentiment-classification-on-stanford-sentiment-treebank-sst-5-test.md?metric=Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| BCN + Char + ELMo | 54.7 | Not recorded | 68fa669a1e209131cea114ca |
| Gumbel Tree-LSTM | 53.7 | Not recorded | 68fa669a1e209131cea114ca |
| Bi-CAS-LSTM | 53.6 | Not recorded | 68fa669a1e209131cea114ca |
| CAS-LSTM | 53 | Not recorded | 68fa669a1e209131cea114ca |
| byte mLSTM | 52.9 | Not recorded | 68fa669a1e209131cea114ca |
| Constituency Tree-LSTM | 52.3 | Not recorded | 68fa669a1e209131cea114ca |
| Constituency Tree-LSTM | 51 | Not recorded | 68fa669a1e209131cea114ca |
| BiLSTM | 48.5 | Not recorded | 68fa669a1e209131cea114ca |
| LSTM | 46 | Not recorded | 68fa669a1e209131cea114ca |
| Recursive Neural Tensor Network | 45.7 | Not recorded | 68fa669a1e209131cea114ca |
| Switch-ACMoE | 40.04 | Not recorded | 68fa77411e209131cea72040 |
| Switch Transformer | 39.13 | Not recorded | 68fa77411e209131cea72040 |
| GLaM-ACMoE | 34.24 | Not recorded | 68fa77411e209131cea72040 |
| Switch-ACMoE | 33.79 | Not recorded | 68fa77411e209131cea72040 |
| GLaM | 33.69 | Not recorded | 68fa77411e209131cea72040 |
| Switch Transformer | 32.21 | Not recorded | 68fa77411e209131cea72040 |

### BCN + Char + ELMo

Accuracy: 54.7. Result ID: 698cfb408b517eb4331e25de.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64271e209131cea02f00.md>)

[Implementation](<http://allennlp.org/elmo>)

Recorded evaluation configuration: {"pre-trained on large external corpora":"true"}. This record does not assert independent reproduction.

### Gumbel Tree-LSTM

Accuracy: 53.7. Result ID: 698cfb408b517eb4331e25dd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63cf1e209131cea00ca0.md>)

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

### Bi-CAS-LSTM

Accuracy: 53.6. Result ID: 698cfb408b517eb4331e25e0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

Recorded evaluation configuration: {"layers":"2"}. This record does not assert independent reproduction.

### CAS-LSTM

Accuracy: 53. Result ID: 698cfb408b517eb4331e25df.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

Recorded evaluation configuration: {"layers":"2"}. This record does not assert independent reproduction.

### byte mLSTM

Accuracy: 52.9. Result ID: 698cfb408b517eb4331e25dc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63371e209131ce9fd805.md>)

Recorded evaluation configuration: {"pre-trained on large external corpora":"true"}. This record does not assert independent reproduction.

### Constituency Tree-LSTM

Accuracy: 52.3. Result ID: 698cfb408b517eb4331e25db.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa61e71e209131ce9f6999.md>)

[Implementation](<https://github.com/stanfordnlp/treelstm>)

Recorded evaluation configuration: {"recurrent dropout":"true"}. This record does not assert independent reproduction.

### Constituency Tree-LSTM

Accuracy: 51. Result ID: 698cfb408b517eb4331e25da.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa61e71e209131ce9f6999.md>)

[Implementation](<https://github.com/stanfordnlp/treelstm>)

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

### BiLSTM

Accuracy: 48.5. Result ID: 698cfb408b517eb4331e25d9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64101e209131cea02761.md>)

Recorded evaluation configuration: {"layers":"2"}. This record does not assert independent reproduction.

### LSTM

Accuracy: 46. Result ID: 698cfb408b517eb4331e25d8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa62a41e209131ce9fa80f.md>)

[Implementation](<https://github.com/stanfordnlp/spinn>)

Recorded evaluation configuration: {"layers":"2"}. This record does not assert independent reproduction.

### Recursive Neural Tensor Network

Accuracy: 45.7. Result ID: 698cfb408b517eb4331e25d7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa669a1e209131cea114ca.md>)

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

### Switch-ACMoE

Accuracy: 40.04. Result ID: 69e7366a47f34ce924cd6ec1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

Recorded evaluation configuration: {"Pre-training dataset":"WikiText-103","Task":"Finetuning"}. This record does not assert independent reproduction.

### Switch Transformer

Accuracy: 39.13. Result ID: 69e7366a47f34ce924cd6ec0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cc11e209131cea35cdf.md>)

Recorded evaluation configuration: {"Pre-training dataset":"WikiText-103","Task":"Finetuning"}. This record does not assert independent reproduction.

### GLaM-ACMoE

Accuracy: 34.24. Result ID: 69e7366a47f34ce924cd6ec3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

Recorded evaluation configuration: {"Pre-training dataset":"WikiText-103","Task":"Finetuning"}. This record does not assert independent reproduction.

### Switch-ACMoE

Accuracy: 33.79. Result ID: 69e7366a47f34ce924cd6ebf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

Recorded evaluation configuration: {"Pre-training dataset":"EnWik-8","Task":"Finetuning"}. This record does not assert independent reproduction.

### GLaM

Accuracy: 33.69. Result ID: 69e7366a47f34ce924cd6ec2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d1d1e209131cea37ddb.md>)

Recorded evaluation configuration: {"Pre-training dataset":"WikiText-103","Task":"Finetuning"}. This record does not assert independent reproduction.

### Switch Transformer

Accuracy: 32.21. Result ID: 69e7366a47f34ce924cd6ebe.

[Reporting paper](<https://www.sota2.com/research/paper/68fa77411e209131cea72040.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cc11e209131cea35cdf.md>)

Recorded evaluation configuration: {"Pre-training dataset":"EnWik-8","Task":"Finetuning"}. This record does not assert independent reproduction.

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