pith:DN3ZR3HK
Gradients with Respect to Semantics Preserving Embeddings Tell the Uncertainty of Large Language Models
Gradients with respect to semantics-preserving embeddings quantify uncertainty in LLM free-form generation.
arxiv:2605.04638 v2 · 2026-05-06 · cs.CL · cs.AI
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{DN3ZR3HKHEQ53TZO2B3X3B24XF}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
We propose the first gradient-based UQ method for free-form generation, SemGrad, which is sampling-free and computationally efficient. ... Experiments demonstrate that both of our methods provide efficient and effective uncertainty estimates, achieving superior performance than state-of-the-art methods, particularly in settings with multiple valid responses.
A confident LLM maintains stable output distributions under semantically equivalent input perturbations, and the Semantic Preservation Score reliably identifies the embeddings that best capture semantics for gradient computation.
SemGrad is a gradient-based uncertainty quantification technique for free-form LLM generation that operates in semantic space using a Semantic Preservation Score to select stable embeddings.
Receipt and verification
| First computed | 2026-06-02T02:04:18.608618Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
1b7798ecea3921ddcf2ed0777d875cb95351abf069d75a5d21ee855cdfb2a073
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/DN3ZR3HKHEQ53TZO2B3X3B24XF \
| 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: 1b7798ecea3921ddcf2ed0777d875cb95351abf069d75a5d21ee855cdfb2a073
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "2ada27bb035079668dcfbf065e13d15f9c1df373dee8c18b021806b2e0a9d7f6",
"cross_cats_sorted": [
"cs.AI"
],
"license": "http://creativecommons.org/licenses/by-nc-sa/4.0/",
"primary_cat": "cs.CL",
"submitted_at": "2026-05-06T08:30:40Z",
"title_canon_sha256": "12cd684de46cdae552f55d885654333db37a55a8636da04b7d1454e861fc95db"
},
"schema_version": "1.0",
"source": {
"id": "2605.04638",
"kind": "arxiv",
"version": 2
}
}