# Emergent Communication Evaluation on ImageNet-100 20-class subset

Source: [Emergent Communication Evaluation on ImageNet-100 20-class subset](<https://www.sota2.com/research/sota/emergent-communication-evaluation-on-imagenet-100-20-class-subset>)

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

Data updated: 2026-04-09T23:40:56.878Z

Evaluation of emergent communication capabilities using the 20-class subset of the ImageNet-100 dataset, reporting Top1 Accuracy, Cosine Similarity, and Overlap metrics.

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 | Emergent Communication Evaluation |
| Dataset (including any stated split) | ImageNet-100 20-class subset |
| Metric | Top-1 Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 8 of 8 |
| 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

- [Top-1 Accuracy](<https://www.sota2.com/research/sota/emergent-communication-evaluation-on-imagenet-100-20-class-subset.md?metric=Top-1%20Accuracy>): Higher is better; unit: PERCENT.

- [Cosine Similarity](<https://www.sota2.com/research/sota/emergent-communication-evaluation-on-imagenet-100-20-class-subset.md?metric=Cosine%20Similarity>): Higher is better; unit: PERCENT.

- [Overlap](<https://www.sota2.com/research/sota/emergent-communication-evaluation-on-imagenet-100-20-class-subset.md?metric=Overlap>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Top-1 Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| SSNG | 44.6 | Not recorded | 68fa756c1e209131cea67cb2 |
| SSNG | 42.91 | Not recorded | 68fa756c1e209131cea67cb2 |
| Reconstruction Game | 24.04 | Not recorded | 68fa756c1e209131cea67cb2 |
| MHNG | 22.8 | Not recorded | 68fa756c1e209131cea67cb2 |
| Referential Game | 19.95 | Not recorded | 68fa756c1e209131cea67cb2 |
| MHNG | 16.47 | Not recorded | 68fa756c1e209131cea67cb2 |
| NoCom | 6.1 | Not recorded | 68fa756c1e209131cea67cb2 |
| NoCom | 5.8 | Not recorded | 68fa756c1e209131cea67cb2 |

### SSNG

Top-1 Accuracy: 44.6. Result ID: 69d838f847f34ce924c00af9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

Recorded evaluation configuration: {"Agent":"Agent A"}. This record does not assert independent reproduction.

### SSNG

Top-1 Accuracy: 42.91. Result ID: 69d838f847f34ce924c00afa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

Recorded evaluation configuration: {"Agent":"Agent B"}. This record does not assert independent reproduction.

### Reconstruction Game

Top-1 Accuracy: 24.04. Result ID: 69d838f847f34ce924c00af4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa675b1e209131cea15cec.md>)

Recorded evaluation configuration: {"Agent":"Speaker"}. This record does not assert independent reproduction.

### MHNG

Top-1 Accuracy: 22.8. Result ID: 69d838f847f34ce924c00af7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e5a1e209131cea3f4db.md>)

Recorded evaluation configuration: {"Agent":"Agent A"}. This record does not assert independent reproduction.

### Referential Game

Top-1 Accuracy: 19.95. Result ID: 69d838f847f34ce924c00af3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63341e209131ce9fd60a.md>)

Recorded evaluation configuration: {"Agent":"Speaker"}. This record does not assert independent reproduction.

### MHNG

Top-1 Accuracy: 16.47. Result ID: 69d838f847f34ce924c00af8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e5a1e209131cea3f4db.md>)

Recorded evaluation configuration: {"Agent":"Agent B"}. This record does not assert independent reproduction.

### NoCom

Top-1 Accuracy: 6.1. Result ID: 69d838f847f34ce924c00af6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

Recorded evaluation configuration: {"Agent":"Agent B"}. This record does not assert independent reproduction.

### NoCom

Top-1 Accuracy: 5.8. Result ID: 69d838f847f34ce924c00af5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa756c1e209131cea67cb2.md>)

Recorded evaluation configuration: {"Agent":"Agent A"}. This record does not assert independent reproduction.

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