What Transfers Under Source Shift? Definitions, Examples, and Fine-Tuning for Climate Disclosure Classification
2026-07-20 • Computation and Language
Computation and LanguageComputational Engineering, Finance, and Science
AI summaryⓘ
The authors studied how well different methods help large language models (LLMs) identify climate-related information across various types of company documents, like reports and press releases. They tested three common adaptation techniques—using definitions, examples, and fine-tuning—on multiple LLMs with datasets from different sources. They found that while all methods improved performance when switching sources, the best methods for one source didn’t always work best for another. Simpler methods, like using matching definitions, tended to work more consistently when the document type changed.
climate disclosurelarge language modelssource shiftcross-source adaptationfew-shot learningfine-tuningsimilarity-based retrievalLoRA fine-tuningtext classificationcorporate reporting
Authors
Guosheng Li, Fenghui Ren, Bin Liu, Chuan Yu, Kaiying Ji, Lin Yue, Jun Shen, Sasa Qian
Abstract
Climate disclosure classification is a fundamental task for analysing corporate climate disclosures, yet such disclosures appear in many different sources -- annual reports, press releases, and earnings calls -- that differ in length, purpose, and writing style. Existing evaluations are mostly conducted within a single source, leaving open whether common LLM adaptation strategies remain effective under source shift. We reframe climate disclosure classification as a cross-source adaptation problem and study three widely used adaptation strategies -- definitions, examples, and fine-tuning -- across eleven open- and closed-source LLMs, using two corpora that share the same label space but come from different sources. We find that all strategies bring positive cross-source gains on average, but the strongest in-source strategies are not the strongest cross-source ones: similarity-based retrieval and LoRA fine-tuning gain most in-source but lose most of that advantage under source shift; randomly selected few-shot examples, a weaker in-source baseline, retain their advantage more reliably; definitions transfer most consistently, though only when their granularity matches the target text. Across these strategies, when the source changes, simpler is often safer.