An automatic debiasing framework rewrites instruction-tuning data to make identified biased features statistically independent of answers, improving zero-shot transfer on NLI and paraphrase tasks.
Mitigate extrinsic social bias in pre-trained language models via continuous prompts adjustment
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Information Gain-Guided Causal Intervention for Autonomous Debiasing Large Language Models
An automatic debiasing framework rewrites instruction-tuning data to make identified biased features statistically independent of answers, improving zero-shot transfer on NLI and paraphrase tasks.