REVIEW 2 cited by
MENLI: Robust Evaluation Metrics from Natural Language Inference
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
read the original abstract
Recently proposed BERT-based evaluation metrics for text generation perform well on standard benchmarks but are vulnerable to adversarial attacks, e.g., relating to information correctness. We argue that this stems (in part) from the fact that they are models of semantic similarity. In contrast, we develop evaluation metrics based on Natural Language Inference (NLI), which we deem a more appropriate modeling. We design a preference-based adversarial attack framework and show that our NLI based metrics are much more robust to the attacks than the recent BERT-based metrics. On standard benchmarks, our NLI based metrics outperform existing summarization metrics, but perform below SOTA MT metrics. However, when combining existing metrics with our NLI metrics, we obtain both higher adversarial robustness (15%-30%) and higher quality metrics as measured on standard benchmarks (+5% to 30%).
Forward citations
Cited by 2 Pith papers
-
Prototypicality Bias Reveals Blindspots in Multimodal Evaluation Metrics
Prototypicality bias: common text-to-image metrics systematically prefer plausible-but-wrong images over correct non-prototypical ones; PROTOSCORE mitigates but does not eliminate the failure.
-
Robust fine-tuning of speech recognition models via model merging: application to disordered speech
Merging multiple fine-tuned Whisper models reduces word error rate on dysarthric speech by 12-16% relative to standard fine-tuning, with gains on long audio and low-data settings.
Discussion (0). Continue with ORCID to comment.