FormatBench evaluates LLM format adherence across ten tasks, and REFF uses format-checker rewards in PPO to raise format faithfulness substantially while keeping content quality roughly stable.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
ReFF: Reinforcing Format Faithfulness in Language Models across Varied Tasks
FormatBench evaluates LLM format adherence across ten tasks, and REFF uses format-checker rewards in PPO to raise format faithfulness substantially while keeping content quality roughly stable.