pith:WIELCL2A
Compliance versus Sensibility: On the Reasoning Controllability in Large Language Models
Large language models prioritize sensible reasoning over following conflicting instructions, but can be steered toward greater compliance.
arxiv:2604.27251 v2 · 2026-04-29 · cs.CL · cs.AI
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{WIELCL2AASD7GFWWWEYLIRVVZJ}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
LLMs consistently prioritize sensibility over compliance, favoring task-appropriate reasoning patterns despite conflicting instructions... we steer models towards compliance, increasing instruction following by up to 29%.
That the constructed reasoning conflicts cleanly isolate parametric versus contextual reasoning without introducing unintended task difficulty or prompt artifacts that could explain the observed sensibility bias.
LLMs favor task-appropriate reasoning over conflicting instructions, yet reasoning types are linearly encoded in middle-to-late layers and can be steered to boost instruction compliance by up to 29%.
Receipt and verification
| First computed | 2026-05-28T01:04:41.094351Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
b208b12f400487f316d6b130b446b5ca7a8092c6138596dbeb73f8f2bc7e489d
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/WIELCL2AASD7GFWWWEYLIRVVZJ \
| 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: b208b12f400487f316d6b130b446b5ca7a8092c6138596dbeb73f8f2bc7e489d
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "9160d6ee20a5c6084a64bb8b41a2930ba9cd30f930792170acb564a259986460",
"cross_cats_sorted": [
"cs.AI"
],
"license": "http://creativecommons.org/licenses/by-nc-sa/4.0/",
"primary_cat": "cs.CL",
"submitted_at": "2026-04-29T22:55:40Z",
"title_canon_sha256": "4eb558136f47d55efc3c7fe1bf1ae844f40e4abde477ef164f23b0ce3056108d"
},
"schema_version": "1.0",
"source": {
"id": "2604.27251",
"kind": "arxiv",
"version": 2
}
}