BatteryPass-12K: The First Dataset for the Novel Digital Battery Passport Conformance Task

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Adewumi, Tosin, Karlsson, Martin, Alkhaled, Lama, Liwicki, Marcus
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866917448192622592
author Adewumi, Tosin
Karlsson, Martin
Alkhaled, Lama
Liwicki, Marcus
author_facet Adewumi, Tosin
Karlsson, Martin
Alkhaled, Lama
Liwicki, Marcus
contents We introduce a novel task of digital battery passport (DBP) conformance classification and introduce the first public benchmark for the task: BatteryPass-12K, created synthetically from real pilot samples. This is as the EU's battery regulation on DBPs comes into effect soon and there exists no public dataset. We evaluated 22 language models (LMs) in zero-shot inference, spanning small LMs (SLMs), mixture of experts (MoEs), and dense LLMs. We also conducted analysis, additional evaluations of few-shot inference and prompt-injection attacks to find that (1) Thinking models have the best performance (with GPT-5.4 scoring 0.98 (0.03) and 0.71 (0.22) on average as F1 (and confidence interval at 95%) on the validation and test sets, respectively), (2) few-shot examples improve performance significantly, (3) generally capable frontier models find the task challenging, (4) merely scaling model parameters does not necessarily lead to improved performance, as SLMs outperformed some LLMs, and (5) prompt-injection attacks degrade performance. We note that BatteryPass-12K, though limited to real pilot samples, may be useful for other known or emerging tasks in the battery domain, e.g. lifecycle reasoning. We publicly release the dataset under a permissive licence (CC-BY-4.0).
format Preprint
id arxiv_https___arxiv_org_abs_2604_26986
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle BatteryPass-12K: The First Dataset for the Novel Digital Battery Passport Conformance Task
Adewumi, Tosin
Karlsson, Martin
Alkhaled, Lama
Liwicki, Marcus
Computation and Language
We introduce a novel task of digital battery passport (DBP) conformance classification and introduce the first public benchmark for the task: BatteryPass-12K, created synthetically from real pilot samples. This is as the EU's battery regulation on DBPs comes into effect soon and there exists no public dataset. We evaluated 22 language models (LMs) in zero-shot inference, spanning small LMs (SLMs), mixture of experts (MoEs), and dense LLMs. We also conducted analysis, additional evaluations of few-shot inference and prompt-injection attacks to find that (1) Thinking models have the best performance (with GPT-5.4 scoring 0.98 (0.03) and 0.71 (0.22) on average as F1 (and confidence interval at 95%) on the validation and test sets, respectively), (2) few-shot examples improve performance significantly, (3) generally capable frontier models find the task challenging, (4) merely scaling model parameters does not necessarily lead to improved performance, as SLMs outperformed some LLMs, and (5) prompt-injection attacks degrade performance. We note that BatteryPass-12K, though limited to real pilot samples, may be useful for other known or emerging tasks in the battery domain, e.g. lifecycle reasoning. We publicly release the dataset under a permissive licence (CC-BY-4.0).
title BatteryPass-12K: The First Dataset for the Novel Digital Battery Passport Conformance Task
topic Computation and Language
url https://arxiv.org/abs/2604.26986