HotpotQA F1 improves across the 56k, 112k and 224k contexts when the maximum action budget increases from 5 to 20, with diminishing gains at larger budgets. Increasing retrieved items from 1 to 5 also improves the plotted F1 curves. · CiteArk
HotpotQA F1 improves across the 56k, 112k and 224k contexts when the maximum action budget increases from 5 to 20, with diminishing gains at larger budgets. Increasing retrieved items from 1 to 5 also improves the plotted F1 curves.
Source: paper:p. 9, Figure 3, §4.5
Reported and observed measurements
No structured measurement is attached to this Claim.
Assessments (0)
No immutable Assessment has been published for this Claim yet.
