Search papers, labs, and topics across Lattice.
RubiCap is introduced as a novel reinforcement learning framework for dense image captioning that leverages LLM-written rubrics to provide fine-grained, sample-specific reward signals. The method assembles a diverse committee of candidate captions, uses an LLM to extract consensus strengths and weaknesses, and converts these insights into explicit evaluation criteria for an LLM judge. Experiments show RubiCap achieves state-of-the-art performance on CapArena, CaptionQA, and improves pretraining of VLMs, even surpassing models trained on proprietary data.
Forget expensive human annotations: RubiCap uses LLM-generated rubrics to train image captioning models via RL, achieving superhuman performance and even improving VLM pretraining.
Dense image captioning is critical for cross-modal alignment in vision-language pretraining and text-to-image generation, but scaling expert-quality annotations is prohibitively expensive. While synthetic captioning via strong vision-language models (VLMs) is a practical alternative, supervised distillation often yields limited output diversity and weak generalization. Reinforcement learning (RL) could overcome these limitations, but its successes have so far been concentrated in verifiable domains that rely on deterministic checkers -- a luxury not available in open-ended captioning. We address this bottleneck with RubiCap, a novel RL framework that derives fine-grained, sample-specific reward signals from LLM-written rubrics. RubiCap first assembles a diverse committee of candidate captions, then employs an LLM rubric writer to extract consensus strengths and diagnose deficiencies in the current policy. These insights are converted into explicit evaluation criteria, enabling an LLM judge to decompose holistic quality assessment and replace coarse scalar rewards with structured, multi-faceted evaluations. Across extensive benchmarks, RubiCap achieves the highest win rates on CapArena, outperforming supervised distillation, prior RL methods, human-expert annotations, and GPT-4V-augmented outputs. On CaptionQA, it demonstrates superior word efficiency: our 7B model matches Qwen2.5-VL-32B-Instruct, and our 3B model surpasses its 7B counterpart. Remarkably, using the compact RubiCap-3B as a captioner produces stronger pretrained VLMs than those trained on captions from proprietary models.