brainsteam.co.uk/brainsteam/content/annotations/2022/11/28/1669635117.md

date	hypothesis-meta	in-reply-to	tags	type	url
2022-11-28T11:31:57	created document flagged group hidden id links permissions tags target text updated uri user user_info 2022-11-28T11:31:57.626263+00:00 title Countering the Effects of Lead Bias in News Summarization via Multi-Stage Training and Auxiliary Losses false __world__ false RG4WUG8QEe2-tZ8v6nNOJA html incontext json https://hypothes.is/a/RG4WUG8QEe2-tZ8v6nNOJA https://hyp.is/RG4WUG8QEe2-tZ8v6nNOJA/aclanthology.org/D19-1620.pdf https://hypothes.is/api/annotations/RG4WUG8QEe2-tZ8v6nNOJA admin delete read update acct:ravenscroftj@hypothes.is acct:ravenscroftj@hypothes.is group:__world__ acct:ravenscroftj@hypothes.is NLProc summarization bandit rl selector source end start type 7922 7890 TextPositionSelector exact prefix suffix type BanditSum a hierarchical bi-LSTM S uses a CNN+bi-GRU encoder, and . RNES’s de-coder is auto-regres TextQuoteSelector https://aclanthology.org/D19-1620.pdf Banditsum uses bi-directional LSTM encoding. It generates sentence-level representations 2022-11-28T11:34:57.447988+00:00 https://aclanthology.org/D19-1620.pdf acct:ravenscroftj@hypothes.is display_name James Ravenscroft	https://aclanthology.org/D19-1620.pdf	NLProc summarization bandit rl hypothesis	annotation	/annotations/2022/11/28/1669635117

BanditSum a hierarchical bi-LSTM

Banditsum uses bi-directional LSTM encoding. It generates sentence-level representations