-
Truth or Mirage? Towards End-to-End Factuality Evaluation with LLM-OASIS
Paper • 2411.19655 • Published • 20 -
Babelscape/LLM-Oasis_claim_extraction
Viewer • Updated • 81.3k • 46 • 6 -
Babelscape/LLM-Oasis_claim_verification
Viewer • Updated • 2.66k • 48 • 5 -
Babelscape/LLM-Oasis_unfactual_text_generation
Viewer • Updated • 81.2k • 19 • 6
AI & ML interests
Babelscape is a deep tech company founded in 2016 focused on multilingual Natural Language Processing.
FENICE is a metric for summarization factuality, with a focus on interpretability. FENICE leverages NLI and claim extraction to assess factuality
-
Babelscape/t5-base-summarization-claim-extractor
0.2B • Updated • 12.3k • 13 -
Babelscape/story-summeval
Viewer • Updated • 319 • 42 • 8 -
Babelscape/FENICE
Updated • 6 -
FENICE: Factuality Evaluation of summarization based on Natural language Inference and Claim Extraction
Paper • 2403.02270 • Published • 3
-
Truth or Mirage? Towards End-to-End Factuality Evaluation with LLM-OASIS
Paper • 2411.19655 • Published • 20 -
Babelscape/LLM-Oasis_claim_extraction
Viewer • Updated • 81.3k • 46 • 6 -
Babelscape/LLM-Oasis_claim_verification
Viewer • Updated • 2.66k • 48 • 5 -
Babelscape/LLM-Oasis_unfactual_text_generation
Viewer • Updated • 81.2k • 19 • 6
Word Sense Linking is the task designed to identify and disambiguate spans of text to their most suitable senses from a reference inventory.
FENICE is a metric for summarization factuality, with a focus on interpretability. FENICE leverages NLI and claim extraction to assess factuality
-
Babelscape/t5-base-summarization-claim-extractor
0.2B • Updated • 12.3k • 13 -
Babelscape/story-summeval
Viewer • Updated • 319 • 42 • 8 -
Babelscape/FENICE
Updated • 6 -
FENICE: Factuality Evaluation of summarization based on Natural language Inference and Claim Extraction
Paper • 2403.02270 • Published • 3