# Procedural Multimodal Machine Comprehension on CraftQA (test)

Source: [Procedural Multimodal Machine Comprehension on CraftQA (test)](<https://www.sota2.com/research/sota/procedural-multimodal-machine-comprehension-on-craftqa-test>)

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

Data updated: 2026-02-26T05:39:25.223Z

Evaluation of Procedural Multimodal Machine Comprehension performance on the test split of the CraftQA dataset, reporting metrics such as Cloze, Coherence, and Ordering scores.

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 | Procedural Multimodal Machine Comprehension |
| Dataset (including any stated split) | CraftQA (test) |
| Metric | Cloze |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| 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

- [Cloze](<https://www.sota2.com/research/sota/procedural-multimodal-machine-comprehension-on-craftqa-test.md?metric=Cloze>): Higher is better; unit: PERCENT.

- [Coherence](<https://www.sota2.com/research/sota/procedural-multimodal-machine-comprehension-on-craftqa-test.md?metric=Coherence>): Higher is better; unit: PERCENT.

- [Ordering](<https://www.sota2.com/research/sota/procedural-multimodal-machine-comprehension-on-craftqa-test.md?metric=Ordering>): Higher is better; unit: PERCENT.

- [Average Score](<https://www.sota2.com/research/sota/procedural-multimodal-machine-comprehension-on-craftqa-test.md?metric=Average%20Score>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Cloze | Result date | Report paper |
| --- | --- | --- | --- |
| Human | 55.33 | Not recorded | 68fa6c2d1e209131cea3290a |
| TMEG | 48.01 | Not recorded | 68fa6c2d1e209131cea3290a |
| TMEG | 47.49 | Not recorded | 68fa6c2d1e209131cea3290a |
| TMEG | 46.67 | Not recorded | 68fa6c2d1e209131cea3290a |
| TMEG | 45.91 | Not recorded | 68fa6c2d1e209131cea3290a |
| TMEG | 44.87 | Not recorded | 68fa6c2d1e209131cea3290a |
| VisualBERT | 40.56 | Not recorded | 68fa6c2d1e209131cea3290a |
| MLMM-Trans | 39.43 | Not recorded | 68fa6c2d1e209131cea3290a |
| PRN | 35.14 | Not recorded | 68fa6c2d1e209131cea3290a |
| HS | 27.81 | Not recorded | 68fa6c2d1e209131cea3290a |

### Human

Cloze: 55.33. Result ID: 698ee9883a6c314849f35923.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66661e209131cea101b3.md>)

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

### TMEG

Cloze: 48.01. Result ID: 698ee9883a6c314849f35928.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

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

### TMEG

Cloze: 47.49. Result ID: 698ee9883a6c314849f3592c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

Recorded evaluation configuration: {"Contrastive Coherence Loss":false}. This record does not assert independent reproduction.

### TMEG

Cloze: 46.67. Result ID: 698ee9883a6c314849f3592a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

Recorded evaluation configuration: {"Modal Encoding":false}. This record does not assert independent reproduction.

### TMEG

Cloze: 45.91. Result ID: 698ee9883a6c314849f35929.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

Recorded evaluation configuration: {"Temporal Encoding":false}. This record does not assert independent reproduction.

### TMEG

Cloze: 44.87. Result ID: 698ee9883a6c314849f3592b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

Recorded evaluation configuration: {"Temporal Encoding":false,"Modal Encoding":false}. This record does not assert independent reproduction.

### VisualBERT

Cloze: 40.56. Result ID: 698ee9883a6c314849f35927.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

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

### MLMM-Trans

Cloze: 39.43. Result ID: 698ee9883a6c314849f35926.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

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

### PRN

Cloze: 35.14. Result ID: 698ee9883a6c314849f35925.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa66661e209131cea101b3.md>)

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

### HS

Cloze: 27.81. Result ID: 698ee9883a6c314849f35924.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2d1e209131cea3290a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa64c81e209131cea06924.md>)

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

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