Applying GRPO reinforcement learning with format and accuracy rewards to a 1.5B LLM yields high weighted F1 on financial tabular benchmarks, but near-zero MCC on imbalanced datasets and unvalidated explanations.
Tabllm: Few-shot classification of tabular data with large language models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
TabReason: A Reinforcement Learning-Enhanced Reasoning LLM for Explainable Tabular Data Prediction
Applying GRPO reinforcement learning with format and accuracy rewards to a 1.5B LLM yields high weighted F1 on financial tabular benchmarks, but near-zero MCC on imbalanced datasets and unvalidated explanations.