Self-play dialog games between two VLMs generate filtered synthetic data that, when fine-tuned on, improves VQA and robotics success detection.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Vision-Language Model Dialog Games for Self-Improvement
Self-play dialog games between two VLMs generate filtered synthetic data that, when fine-tuned on, improves VQA and robotics success detection.