pith:VEH5UWC6
Why are language models less surprised than humans? Testing the Parse Multiplicity Mismatch Hypothesis
Reducing the number of simultaneous parses in language models increases predicted garden path effects but not enough to match human reading difficulties.
arxiv:2605.15440 v1 · 2026-05-14 · cs.CL
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{VEH5UWC6AZYXWPFKHB55FIIQWV}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
Reducing the number of simultaneous active parses indeed increases the magnitude of predicted garden path effects, but not nearly enough to capture the full magnitude of the effects in humans.
That word-synchronous beam search in RNNGs with a controllable beam size provides a faithful proxy for the number of distinct sentence interpretations a human parser can maintain in parallel (section on methods and results).
Varying the number of simultaneous parses in RNNGs increases predicted garden-path effects but does not fully reconcile LM surprisal with human reading times.
References
Formal links
Receipt and verification
| First computed | 2026-05-20T00:00:58.672446Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
a90fda585e06717b3caa387bd2a110b579d75dcc46fb86728b06407dc5d15cc5
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/VEH5UWC6AZYXWPFKHB55FIIQWV \
| jq -c '.canonical_record' \
| python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: a90fda585e06717b3caa387bd2a110b579d75dcc46fb86728b06407dc5d15cc5
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "aa038da97d8ee8e0569cb058715aad053020e327d658c3ad6505862823d22575",
"cross_cats_sorted": [],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "cs.CL",
"submitted_at": "2026-05-14T21:44:26Z",
"title_canon_sha256": "164ba0f468b1cc8e9a35dc7629451a19dfcb5457957a3fc03856cd45488c6856"
},
"schema_version": "1.0",
"source": {
"id": "2605.15440",
"kind": "arxiv",
"version": 1
}
}