pith:SPXEBG6S
Skill Neologisms: Towards Skill-based Continual Learning
Skill neologisms let LLMs gain new abilities by adding optimized soft tokens to the vocabulary without any weight updates.
arxiv:2605.04970 v2 · 2026-05-06 · cs.LG · cs.AI
Add to your LaTeX paper
\usepackage{pith}
\pithnumber{SPXEBG6S3HEMQMNL4342VK2377}
Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge
Record completeness
Claims
We then show that skill neologisms can be learned to improve model capabilities on specific skills while being composable with out-of-distribution skills, and that independently trained skill neologisms can be composed zero-shot. These results suggest that skill neologisms may provide a scalable path towards skill-based continual learning.
That independently optimized skill neologisms can be added and composed without degrading the base model's performance on unrelated tasks or causing interference between skills.
Skill neologisms are optimized soft tokens that improve LLM performance on targeted skills without weight updates and allow zero-shot composition for continual learning.
Receipt and verification
| First computed | 2026-05-20T02:05:44.317585Z |
|---|---|
| Builder | pith-number-builder-2026-05-17-v1 |
| Signature | Pith Ed25519
(pith-v1-2026-05) · public key |
| Schema | pith-number/v1.0 |
Canonical hash
93ee409bd2d9c8c831abe6f9aaab5bffc0b50083bfae51650f149a540bcdc680
Aliases
· · · · ·Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/SPXEBG6S3HEMQMNL4342VK2377 \
| 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: 93ee409bd2d9c8c831abe6f9aaab5bffc0b50083bfae51650f149a540bcdc680
Canonical record JSON
{
"metadata": {
"abstract_canon_sha256": "20a2ba7a276103402d440f2cc86dd85696c2dca5ff913c3cc6490f490b843dea",
"cross_cats_sorted": [
"cs.AI"
],
"license": "http://creativecommons.org/licenses/by/4.0/",
"primary_cat": "cs.LG",
"submitted_at": "2026-05-06T14:27:12Z",
"title_canon_sha256": "eaf65003074448d82097cf0cb8bfcad3c9dfe2c0443a06e3a9096a73c1e91a5e"
},
"schema_version": "1.0",
"source": {
"id": "2605.04970",
"kind": "arxiv",
"version": 2
}
}