Fine-tuned LLaMA-3.1-8B generates valid, plausible counterfactuals for sensor-based stress prediction, and augmenting with them recovers ~20% of F1 under label scarcity.
Model-agnostic counterfactual explanations for consequential decisions
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Counterfactual Modeling with Fine-Tuned LLMs for Health Intervention Design and Sensor Data Augmentation
Fine-tuned LLaMA-3.1-8B generates valid, plausible counterfactuals for sensor-based stress prediction, and augmenting with them recovers ~20% of F1 under label scarcity.