# Twitter Spam dataset (test)

Source: [Twitter Spam dataset (test)](<https://www.sota2.com/research/sota/twitter-spam-dataset-test>)

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

Data updated: 2026-02-26T05:39:15.308Z

Evaluation of Spam Detection performance on the test split of the Twitter Spam dataset.

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 | Spam Detection |
| Dataset (including any stated split) | Twitter Spam dataset (test) |
| Metric | Clean Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 26 of 26 |
| 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

- [Clean Accuracy](<https://www.sota2.com/research/sota/twitter-spam-dataset-test.md?metric=Clean%20Accuracy>): Higher is better; unit: PERCENT.

- [Defense Success Rate](<https://www.sota2.com/research/sota/twitter-spam-dataset-test.md?metric=Defense%20Success%20Rate>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Clean Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| Std. Training | 94.6 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-Pearson | 94 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-SHAP | 93.9 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MD | 93.9 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MDtarget | 93.9 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| Uniform-δ | 92.6 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MDtarget | 92 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-SHAP | 91.9 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MD | 91.9 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-Pearson | 91.8 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| Uniform-δ | 91.3 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-Pearson | 91.2 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MDtarget | 91.2 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| Uniform-δ | 91.1 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MD | 91.1 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| Uniform-δ | 91.1 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-SHAP | 90.9 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-Pearson | 90.6 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-SHAP | 90.5 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MD | 90.5 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MDtarget | 90.5 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| Uniform-δ | 90.5 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-Pearson | 89.9 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-SHAP | 89.8 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MDtarget | 89.8 | Not recorded | 68fa6b2c1e209131cea2c8a7 |
| NU-δ-MD | 89.7 | Not recorded | 68fa6b2c1e209131cea2c8a7 |

### Std. Training

Clean Accuracy: 94.6. Result ID: 698e65ecffc3901de09d40bb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

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

### NU-δ-Pearson

Clean Accuracy: 94. Result ID: 698e65ecffc3901de09d40be.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.1}. This record does not assert independent reproduction.

### NU-δ-SHAP

Clean Accuracy: 93.9. Result ID: 698e65ecffc3901de09d40bd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.1}. This record does not assert independent reproduction.

### NU-δ-MD

Clean Accuracy: 93.9. Result ID: 698e65ecffc3901de09d40bf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.1}. This record does not assert independent reproduction.

### NU-δ-MDtarget

Clean Accuracy: 93.9. Result ID: 698e65ecffc3901de09d40c0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.1}. This record does not assert independent reproduction.

### Uniform-δ

Clean Accuracy: 92.6. Result ID: 698e65ecffc3901de09d40c1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.3}. This record does not assert independent reproduction.

### NU-δ-MDtarget

Clean Accuracy: 92. Result ID: 698e65ecffc3901de09d40c5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.3}. This record does not assert independent reproduction.

### NU-δ-SHAP

Clean Accuracy: 91.9. Result ID: 698e65ecffc3901de09d40c2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.3}. This record does not assert independent reproduction.

### NU-δ-MD

Clean Accuracy: 91.9. Result ID: 698e65ecffc3901de09d40c4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.3}. This record does not assert independent reproduction.

### NU-δ-Pearson

Clean Accuracy: 91.8. Result ID: 698e65ecffc3901de09d40c3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.3}. This record does not assert independent reproduction.

### Uniform-δ

Clean Accuracy: 91.3. Result ID: 698e65ecffc3901de09d40c6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.5}. This record does not assert independent reproduction.

### NU-δ-Pearson

Clean Accuracy: 91.2. Result ID: 698e65ecffc3901de09d40c8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.5}. This record does not assert independent reproduction.

### NU-δ-MDtarget

Clean Accuracy: 91.2. Result ID: 698e65ecffc3901de09d40ca.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.5}. This record does not assert independent reproduction.

### Uniform-δ

Clean Accuracy: 91.1. Result ID: 698e65ecffc3901de09d40bc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.1}. This record does not assert independent reproduction.

### NU-δ-MD

Clean Accuracy: 91.1. Result ID: 698e65ecffc3901de09d40c9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.5}. This record does not assert independent reproduction.

### Uniform-δ

Clean Accuracy: 91.1. Result ID: 698e65ecffc3901de09d40cb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.7}. This record does not assert independent reproduction.

### NU-δ-SHAP

Clean Accuracy: 90.9. Result ID: 698e65ecffc3901de09d40c7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.5}. This record does not assert independent reproduction.

### NU-δ-Pearson

Clean Accuracy: 90.6. Result ID: 698e65ecffc3901de09d40cd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.7}. This record does not assert independent reproduction.

### NU-δ-SHAP

Clean Accuracy: 90.5. Result ID: 698e65ecffc3901de09d40cc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.7}. This record does not assert independent reproduction.

### NU-δ-MD

Clean Accuracy: 90.5. Result ID: 698e65ecffc3901de09d40ce.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.7}. This record does not assert independent reproduction.

### NU-δ-MDtarget

Clean Accuracy: 90.5. Result ID: 698e65ecffc3901de09d40cf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":0.7}. This record does not assert independent reproduction.

### Uniform-δ

Clean Accuracy: 90.5. Result ID: 698e65ecffc3901de09d40d0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":1}. This record does not assert independent reproduction.

### NU-δ-Pearson

Clean Accuracy: 89.9. Result ID: 698e65ecffc3901de09d40d2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":1}. This record does not assert independent reproduction.

### NU-δ-SHAP

Clean Accuracy: 89.8. Result ID: 698e65ecffc3901de09d40d1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":1}. This record does not assert independent reproduction.

### NU-δ-MDtarget

Clean Accuracy: 89.8. Result ID: 698e65ecffc3901de09d40d4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":1}. This record does not assert independent reproduction.

### NU-δ-MD

Clean Accuracy: 89.7. Result ID: 698e65ecffc3901de09d40d3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b2c1e209131cea2c8a7.md>)

[Implementation](<https://github.com/amazon-research/adversarial-robustness-with-nonuniform-perturbations>)

Recorded evaluation configuration: {"\|\|δ\|\|2":1}. This record does not assert independent reproduction.

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