analiza aiqbench
TabSieve addresses the brittleness of current LLM-based tabular prediction by introducing an explicit "select-then-predict" framework that audibly identifies informative rows as evidence. The method combines synthetic data training with a reinforcement learning recipe to jointly optimize evidence selection and prediction accuracy, demonstrating robustness against noisy context in both classification and regression tasks. This work advances the field of machine learning on structured data by making the reasoning process more transparent and reliable.
TabSieve: Explicit In-Table Evidence Selection for Tabular Prediction
arXiv:2602.11700v2 Announce Type: replace Abstract: Tabular prediction can benefit from in-table rows as few-shot evidence, yet existing tabular models typically perform instance-wise inference and LLM-based prompting is often brittle. Models do not consistently leverage relevant rows, and noisy context can degrade performance. To address this challenge, we propose TabSieve, a select-then-predict framework that makes evidence usage explicit and auditable. Given a table and a query row, TabSieve first selects a small set of informative rows as evidence and then predicts the missing target conditioned on the selected evidence. To enable this capability, we construct TabSieve-SFT-40K by synthesizing high-quality reasoning trajectories from 331 real tables using a strong teacher model with strict filtering. Furthermore, we introduce TAB-GRPO, a reinforcement learning recipe that jointly optimizes evidence selection and prediction correctness with separate rewards, and stabilizes mixed regression and classification training via dynamic task-advantage balancing. Experiments on a held-out benchmark of 75 classification and 52 regression tables show that TabSieve consistently improves performance across shot budgets, with average gains of 2.92% on classification and 4.45% on regression over the second-best baseline. Further analysis indicates that TabSieve concentrates more attention on the selected evidence, which improves robustness to noisy context.
Abstract:Tabular prediction can benefit from in-table rows as few-shot evidence, yet existing tabular models typically perform instance-wise inference and LLM-based prompting is often brittle. Models do not consistently leverage relevant rows, and noisy context can degrade performance. To address this challenge, we propose TabSieve, a select-then-predict framework that makes evidence usage explicit and auditable. Given a table and a query row, TabSieve first selects a small set of informative rows as evidence and then predicts the missing target conditioned on the selected evidence. To enable this capability, we construct TabSieve-SFT-40K by synthesizing high-quality reasoning trajectories from 331 real tables using a strong teacher model with strict filtering. Furthermore, we introduce TAB-GRPO, a reinforcement learning recipe that jointly optimizes evidence selection and prediction correctness with separate rewards, and stabilizes mixed regression and classification training via dynamic task-advantage balancing. Experiments on a held-out benchmark of 75 classification and 52 regression tables show that TabSieve consistently improves performance across shot budgets, with average gains of 2.92% on classification and 4.45% on regression over the second-best baseline. Further analysis indicates that TabSieve concentrates more attention on the selected evidence, which improves robustness to noisy context.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2602.11700 [cs.LG]
(or arXiv:2602.11700v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2602.11700 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 286 cuvinte
Subiecte conexe
The Entangled Dimensions of Decision Theory
LessWrong · 30 iul. 2026
Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident
METR · 26 aug. 2026
Always ask your agent to be honest
LessWrong · 16 sept. 2026
Reflections on unlearning and inoculation
LessWrong · 20 sept. 2026