Pith Number

pith:FHPN6PIU

pith:2025:FHPN6PIUTIZ72DNJ6WAUUQRAK6

not attested not anchored not stored refs resolved

PEFT-Bench: A Parameter-Efficient Fine-Tuning Methods Benchmark

Branislav Pecher, Ivan Srba, Maria Bielikova, Robert Belanec

PEFT-Bench offers a standardized way to compare parameter-efficient fine-tuning methods for large language models while factoring in training and inference costs.

arxiv:2511.21285 v3 · 2025-11-26 · cs.CL

Open paper page JSON What is a Pith Number?

Record completeness

1 Bitcoin timestamp

2 Internet Archive

3 Author claim open · sign in to claim

4 Citations open

5 Replications open

Claims

C1strongest claim

We introduce PEFT-Bench, a unified end-to-end benchmark for evaluating diverse PEFT methods on autoregressive LLMs. We demonstrate its usage across 27 NLP datasets and 7 PEFT methods. To account for different PEFT training and inference factors, we also introduce the PEFT Soft Cost Penalties (PSCP) metric, which takes trainable parameters, inference speed, and training memory usage into account.

C2weakest assumption

That the chosen 27 NLP datasets and 7 PEFT methods form a sufficiently representative sample to support general conclusions about PEFT method quality and that the PSCP weighting of cost factors produces practically useful rankings.

C3one line summary

PEFT-Bench is a standardized end-to-end benchmark for 7 PEFT methods across 27 NLP datasets on autoregressive LLMs, accompanied by the PSCP metric that penalizes based on trainable parameters, inference speed, and training memory.

References

71 extracted · 71 resolved · 12 Pith anchors

[1] online" 'onlinestring :=

[2] write newline

[3] GPT-4 Technical Report 2023 · arXiv:2303.08774

[4] M ath QA : Towards interpretable math word problem solving with operation-based formalisms 2019 · doi:10.18653/v1/n19-1245

[5] Akari Asai, Mohammadreza Salehi, Matthew Peters, and Hannaneh Hajishirzi. 2022. https://doi.org/10.18653/v1/2022.emnlp-main.446 ATTEMPT : Parameter-efficient multi-task tuning via attentional mixtures 2022 · doi:10.18653/v1/2022.emnlp-main.446

Formal links

1 machine-checked theorem link

Cited by

2 papers in Pith

PEFT-Factory: Unified Parameter-Efficient Fine-Tuning of Autoregressive Large Language Models

Low-Data Supervised Adaptation Outperforms Prompting for Cloud Segmentation Under Domain Shift

Receipt and verification

First computed	2026-05-18T03:10:11.765326Z
Builder	pith-number-builder-2026-05-17-v1
Signature	Pith Ed25519 (`pith-v1-2026-05`) · public key
Schema	pith-number/v1.0

Canonical hash

29dedf3d149a33fd0da9f5814a422057bf4df4d6cc9a502e4cd9cca979605b08

Aliases

arxiv: 2511.21285 · arxiv_version: 2511.21285v3 · doi: 10.48550/arxiv.2511.21285

Agent API

Resolver JSON Graph JSON Events JSON Schema Signing key

Verify this Pith Number yourself

curl -sH 'Accept: application/ld+json' https://pith.science/pith/FHPN6PIUTIZ72DNJ6WAUUQRAK6 \
  | 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: 29dedf3d149a33fd0da9f5814a422057bf4df4d6cc9a502e4cd9cca979605b08

Canonical record JSON

{
  "metadata": {
    "abstract_canon_sha256": "061f942babfc7a00a8c10427f503abb2c2932895185c7def1112ace880ad0707",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.CL",
    "submitted_at": "2025-11-26T11:18:06Z",
    "title_canon_sha256": "c89f7bef98f8708346256ccd3fac07f49c8b339380cc8a53a8b22757b620fc58"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2511.21285",
    "kind": "arxiv",
    "version": 3
  }
}