REVIEW 1 cited by
The Accuracy, Robustness, and Readability of LLM-Generated Sustainability-Related Word Definitions
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
abstract
A common language with standardized definitions is crucial for effective climate discussions. However, concerns exist about LLMs misrepresenting climate terms. We compared 300 official IPCC glossary definitions with those generated by GPT-4o-mini, Llama3.1 8B, and Mistral 7B, analyzing adherence, robustness, and readability using SBERT sentence embeddings. The LLMs scored an average adherence of $0.57-0.59 \pm 0.15$, and their definitions proved harder to read than the originals. Model-generated definitions vary mainly among words with multiple or ambiguous definitions, showing the potential to highlight terms that need standardization. The results show how LLMs could support environmental discourse while emphasizing the need to align model outputs with established terminology for clarity and consistency.
Forward citations
Cited by 1 Pith paper
-
EarthSE: A Benchmark for Evaluating Earth Scientific Exploration Capability of LLMs
EarthSE provides a two-level QA benchmark and an open-ended dialogue benchmark for Earth science and shows current LLMs perform poorly on both.
Discussion (0). Continue with ORCID to comment.