Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

3D-LLM

Benchmarks

Task NameDataset NameSOTA ResultTrend
3D Task Decomposition3D-LLM Task Decomposition
Accuracy52.2
8
3D Dialogue3D-LLM 3D-assisted Dialogue
Accuracy70.6
8
Task Decomposition3D-LLM Held-In 1.0 (test)
BLEU-143.8
6
3D-assisted Dialog3D-LLM Held-In Dataset 1.0 (test)
BLEU-147.4
6
Showing 4 of 4 rows