Generally, the percent of ionic character in a two-element compound correlates quite well with the difference in the electronegativities of the two elements making up the compound, as can be seen in this graph:
(Image originally by the University of Florida, via user4076 here at ChemSE.)
Even here, though, an anomaly is visible.
- HF has an electronegativity difference of 1.893 (4.193 - 2.300).
- LiBr has an electronegativity difference of 1.773 (2.685 - 0.912).
- KI has an electronegativity difference of 1.625 (2.359 - 0.734).
- LiI has an electronegativity difference of 1.447 (2.359 - 0.912).
Based on this, HF should be mostly ionic - certainly more so than LiBr, KI, or especially LiI. Instead, judging from the graph, HF is slightly more covalent than ionic, with ~45% ionic character, whereas LiBr is ~60% ionic, KI is ~75% ionic, and even LiI is ~52-53% ionic despite having a considerably-lower electronegativity difference than the ~45%-ionic HF.
Going along the sequence KI - LiBr - HF, the ionic character of the compound actually declines steeply as the electronegativity difference gets somewhat greater.
What is going on here?
