brainsteam.co.uk/brainsteam/content/annotations/2022/11/23/1669236944.md

date	hypothesis-meta	in-reply-to	tags	type	url
2022-11-23T20:55:44	created document flagged group hidden id links permissions tags target text updated uri user user_info 2022-11-23T20:55:44.414977+00:00 title 2022.naacl-main.167.pdf false __world__ false MrGLumtxEe21b1OADBLmyg html incontext json https://hypothes.is/a/MrGLumtxEe21b1OADBLmyg https://hyp.is/MrGLumtxEe21b1OADBLmyg/aclanthology.org/2022.naacl-main.167.pdf https://hypothes.is/api/annotations/MrGLumtxEe21b1OADBLmyg admin delete read update acct:ravenscroftj@hypothes.is acct:ravenscroftj@hypothes.is group:__world__ acct:ravenscroftj@hypothes.is prompt-models NLProc selector source end start type 20146 19539 TextPositionSelector exact prefix suffix type Misleading Templates There is no consistent re-lation between the performance of models trainedwith templates that are moderately misleading (e.g.{premise} Can that be paraphrasedas "{hypothesis}"?) vs. templates that areextremely misleading (e.g., {premise} Isthis a sports news? {hypothesis}).T0 (both 3B and 11B) perform better givenmisleading-moderate (Figure 3), ALBERT andT5 3B perform better given misleading-extreme(Appendices E and G.4), whereas T5 11B andGPT-3 perform comparably on both sets (Figure 2;also see Table 2 for a summary of statisticalsignificances.) Despite a lack of pattern between structiveand misleading-extreme. 4 8 16 32 64 128 2560.50.550.60. TextQuoteSelector https://aclanthology.org/2022.naacl-main.167.pdf Their misleading templates really are misleading {premise} Can that be paraphrased as "{hypothesis}" {premise} Is this a sports news? {hypothesis} 2022-11-23T20:55:44.414977+00:00 https://aclanthology.org/2022.naacl-main.167.pdf acct:ravenscroftj@hypothes.is display_name James Ravenscroft	https://aclanthology.org/2022.naacl-main.167.pdf	prompt-models NLProc hypothesis	annotation	/annotation/2022/11/23/1669236944

Misleading Templates There is no consistent re-lation between the performance of models trainedwith templates that are moderately misleading (e.g.{premise} Can that be paraphrasedas "{hypothesis}"?) vs. templates that areextremely misleading (e.g., {premise} Isthis a sports news? {hypothesis}).T0 (both 3B and 11B) perform better givenmisleading-moderate (Figure 3), ALBERT andT5 3B perform better given misleading-extreme(Appendices E and G.4), whereas T5 11B andGPT-3 perform comparably on both sets (Figure 2;also see Table 2 for a summary of statisticalsignificances.) Despite a lack of pattern between

Their misleading templates really are misleading

{premise} Can that be paraphrased as "{hypothesis}"

{premise} Is this a sports news? {hypothesis}