Consensus Entropy measures inter-VLM output agreement to verify OCR reliability and enable self-improving ensembles, yielding 42.1% F1 gains over single-model judging.
Layoutlmv3: Pre-training for document ai with unified text and image masking
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CV 2verdicts
CONDITIONAL 2roles
background 1polarities
background 1representative citing papers
A decomposition-first pipeline with topology-aware chunking and interface-constrained merging converts full clinical guidelines into executable decision graphs, raising edge precision from 19.6% to 69.0% and triplet recall from 16.1% to 87.5% on a prostate guideline benchmark.
citing papers explorer
-
Consensus Entropy: Harnessing Multi-VLM Agreement for Self-Verifying and Self-Improving OCR
Consensus Entropy measures inter-VLM output agreement to verify OCR reliability and enable self-improving ensembles, yielding 42.1% F1 gains over single-model judging.
-
Guideline2Graph: Profile-Aware Multimodal Parsing for Executable Clinical Decision Graphs
A decomposition-first pipeline with topology-aware chunking and interface-constrained merging converts full clinical guidelines into executable decision graphs, raising edge precision from 19.6% to 69.0% and triplet recall from 16.1% to 87.5% on a prostate guideline benchmark.