# 3D-Text Retrieval on Objaverse-LVIS 1.0 (test)

Source: [3D-Text Retrieval on Objaverse-LVIS 1.0 (test)](<https://www.sota2.com/research/sota/3d-text-retrieval-on-objaverse-lvis-1-0-test>)

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

Data updated: 2026-02-26T05:38:33.696Z

Evaluation of 3D-Text Retrieval performance using the test split of the Objaverse-LVIS 1.0 dataset, assessed using language model-based retrieval metrics (e.g., CLIP, RoBERTa, BERT Top-5).

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 | 3D-Text Retrieval |
| Dataset (including any stated split) | Objaverse-LVIS 1.0 (test) |
| Metric | CLIP Top-5 Retrieval |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 15 of 15 |
| 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

- [CLIP Top-5 Retrieval](<https://www.sota2.com/research/sota/3d-text-retrieval-on-objaverse-lvis-1-0-test.md?metric=CLIP%20Top-5%20Retrieval>): Higher is better; unit: PERCENT.

- [RoBERTa Top-5 Retrieval](<https://www.sota2.com/research/sota/3d-text-retrieval-on-objaverse-lvis-1-0-test.md?metric=RoBERTa%20Top-5%20Retrieval>): Higher is better; unit: PERCENT.

- [BERT Top-5 Retrieval](<https://www.sota2.com/research/sota/3d-text-retrieval-on-objaverse-lvis-1-0-test.md?metric=BERT%20Top-5%20Retrieval>): Higher is better; unit: PERCENT.

## Recorded results

| Method | CLIP Top-5 Retrieval | Result date | Report paper |
| --- | --- | --- | --- |
| Affine + Subspace Projection | 85.6 | Not recorded | 68fa790f1e209131cea7be7e |
| Affine + Subspace Projection | 85.2 | Not recorded | 68fa790f1e209131cea7be7e |
| Affine + Subspace Projection | 81.6 | Not recorded | 68fa790f1e209131cea7be7e |
| Local CKA + Subspace Projection | 60.19 | Not recorded | 68fa790f1e209131cea7be7e |
| Local CKA + Subspace Projection | 56.9 | Not recorded | 68fa790f1e209131cea7be7e |
| Local CKA + Subspace Projection | 53 | Not recorded | 68fa790f1e209131cea7be7e |
| Affine + Subspace Projection | 45.6 | Not recorded | 68fa790f1e209131cea7be7e |
| Affine + Subspace Projection | 42.2 | Not recorded | 68fa790f1e209131cea7be7e |
| Affine + Subspace Projection | 36.6 | Not recorded | 68fa790f1e209131cea7be7e |
| Affine | 34.4 | Not recorded | 68fa790f1e209131cea7be7e |
| Affine | 23.6 | Not recorded | 68fa790f1e209131cea7be7e |
| Affine | 21.8 | Not recorded | 68fa790f1e209131cea7be7e |
| Local CKA | 18 | Not recorded | 68fa790f1e209131cea7be7e |
| Local CKA | 15.2 | Not recorded | 68fa790f1e209131cea7be7e |
| Local CKA | 13.6 | Not recorded | 68fa790f1e209131cea7be7e |

### Affine + Subspace Projection

CLIP Top-5 Retrieval: 85.6. Result ID: 6995c18c3badfaffe6c424f5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"OpenShape","3D Encoder Category":"Multi-modal"}. This record does not assert independent reproduction.

### Affine + Subspace Projection

CLIP Top-5 Retrieval: 85.2. Result ID: 6995c18c3badfaffe6c424f6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"ULIP-2","3D Encoder Category":"Multi-modal"}. This record does not assert independent reproduction.

### Affine + Subspace Projection

CLIP Top-5 Retrieval: 81.6. Result ID: 6995c18c3badfaffe6c424f7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"Uni3D","3D Encoder Category":"Multi-modal"}. This record does not assert independent reproduction.

### Local CKA + Subspace Projection

CLIP Top-5 Retrieval: 60.19. Result ID: 6995c18c3badfaffe6c42501.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"PointBert","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Local CKA + Subspace Projection

CLIP Top-5 Retrieval: 56.9. Result ID: 6995c18c3badfaffe6c42502.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"SparseConv","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Local CKA + Subspace Projection

CLIP Top-5 Retrieval: 53. Result ID: 6995c18c3badfaffe6c42503.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"Pointnet","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Affine + Subspace Projection

CLIP Top-5 Retrieval: 45.6. Result ID: 6995c18c3badfaffe6c424fc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"SparseConv","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Affine + Subspace Projection

CLIP Top-5 Retrieval: 42.2. Result ID: 6995c18c3badfaffe6c424fb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"PointBert","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Affine + Subspace Projection

CLIP Top-5 Retrieval: 36.6. Result ID: 6995c18c3badfaffe6c424fd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Implementation](<https://github.com/Souhail-01/3d-text-alignment>)

Recorded evaluation configuration: {"3D Encoder":"Pointnet","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Affine

CLIP Top-5 Retrieval: 34.4. Result ID: 6995c18c3badfaffe6c424f9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa722a1e209131cea54e66.md>)

[Implementation](<https://github.com/Flegyas/latent-translation>)

Recorded evaluation configuration: {"3D Encoder":"SparseConv","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Affine

CLIP Top-5 Retrieval: 23.6. Result ID: 6995c18c3badfaffe6c424f8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa722a1e209131cea54e66.md>)

[Implementation](<https://github.com/Flegyas/latent-translation>)

Recorded evaluation configuration: {"3D Encoder":"PointBert","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Affine

CLIP Top-5 Retrieval: 21.8. Result ID: 6995c18c3badfaffe6c424fa.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa722a1e209131cea54e66.md>)

[Implementation](<https://github.com/Flegyas/latent-translation>)

Recorded evaluation configuration: {"3D Encoder":"Pointnet","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Local CKA

CLIP Top-5 Retrieval: 18. Result ID: 6995c18c3badfaffe6c42500.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72a31e209131cea57ba1.md>)

[Implementation](<https://www.sota2.com/github.com/mayug/0-shot-llm-vision>)

Recorded evaluation configuration: {"3D Encoder":"Pointnet","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Local CKA

CLIP Top-5 Retrieval: 15.2. Result ID: 6995c18c3badfaffe6c424fe.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72a31e209131cea57ba1.md>)

[Implementation](<https://www.sota2.com/github.com/mayug/0-shot-llm-vision>)

Recorded evaluation configuration: {"3D Encoder":"PointBert","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

### Local CKA

CLIP Top-5 Retrieval: 13.6. Result ID: 6995c18c3badfaffe6c424ff.

[Reporting paper](<https://www.sota2.com/research/paper/68fa790f1e209131cea7be7e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72a31e209131cea57ba1.md>)

[Implementation](<https://www.sota2.com/github.com/mayug/0-shot-llm-vision>)

Recorded evaluation configuration: {"3D Encoder":"SparseConv","3D Encoder Category":"Uni-modal"}. This record does not assert independent reproduction.

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