The paper proposes MQG-RFM, which fine-tunes a patent retriever on LLM-generated query variations, and reports dramatic accuracy gains that are undercut by uncontrolled baseline comparisons and unverifiable numbers.
utility models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Retrieval-Augmented Generation Systems for Intellectual Property via Synthetic Multi-Angle Fine-tuning
The paper proposes MQG-RFM, which fine-tunes a patent retriever on LLM-generated query variations, and reports dramatic accuracy gains that are undercut by uncontrolled baseline comparisons and unverifiable numbers.