Varying the number of simultaneous parses in RNNGs increases predicted garden-path effects but does not fully reconcile LM surprisal with human reading times.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
CONF-LA delivers a low-latency online method for line assignment in noisy gaze data that narrows the online-offline performance gap to 1-2% and reaches ~95% median accuracy on children's data.
Early layers of language models predict early-pass human reading times better than surprisal, with surprisal superior for late-pass measures and strong variation by language.
citing papers explorer
-
Why are language models less surprised than humans? Testing the Parse Multiplicity Mismatch Hypothesis
Varying the number of simultaneous parses in RNNGs increases predicted garden-path effects but does not fully reconcile LM surprisal with human reading times.
-
Sure About That Line? Approaching Confidence-Based, Real-Time Line Assignment in Reading Gaze Data
CONF-LA delivers a low-latency online method for line assignment in noisy gaze data that narrows the online-offline performance gap to 1-2% and reaches ~95% median accuracy on children's data.
-
Probing for Reading Times
Early layers of language models predict early-pass human reading times better than surprisal, with surprisal superior for late-pass measures and strong variation by language.