# Polaris: Learning to Generate Table Descriptions from Retrieval Feedback

arXiv: https://arxiv.org/abs/2608.17171 · HTML: https://mlsanity.com/paper/2608.17171/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Preference alignment for recommenders and RAG, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 29 of its 49 references that are in the map.
- First posted on arXiv 17 Aug 2026.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Preference alignment for recommenders and RAG](https://mlsanity.com/alignment/preference-alignment-for-recommenders-and-rag/)

## Builds on (most cited first, 29 of 29)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Dense Passage Retrieval for Open-Domain Question Answering](https://mlsanity.com/paper/2004.04906/) (2020)
- [QLoRA: Efficient Finetuning of Quantized LLMs](https://mlsanity.com/paper/2305.14314/) (2023)
- [Fine-Tuning Language Models from Human Preferences](https://mlsanity.com/paper/1909.08593/) (2019)
- [RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback](https://mlsanity.com/paper/2309.00267/) (2023)
- [Zephyr: Direct Distillation of LM Alignment](https://mlsanity.com/paper/2310.16944/) (2023)
- [SPLADE: Sparse Lexical and Expansion Model for First Stage Ranking](https://mlsanity.com/paper/2107.05720/) (2021)
- [Sherlock: A Deep Learning Approach to Semantic Data Type Detection](https://mlsanity.com/paper/1905.10688/) (2019)
- [An Analysis of Fusion Functions for Hybrid Retrieval](https://mlsanity.com/paper/2210.11934/) (2022)
- [Valentine: Evaluating Matching Techniques for Dataset Discovery](https://mlsanity.com/paper/2010.07386/) (2020)
- [ReasTAP: Injecting Table Reasoning Skills During Pre-training via Synthetic Reasoning Examples](https://mlsanity.com/paper/2210.12374/) (2022)
- [StruBERT: Structure-aware BERT for Table Search and Matching](https://mlsanity.com/paper/2203.14278/) (2022)
- [CHORUS: Foundation Models for Unified Data Discovery and Exploration](https://mlsanity.com/paper/2306.09610/) (2023)
- [Pneuma: Leveraging LLMs for Tabular Data Representation and Retrieval in an End-to-End System](https://mlsanity.com/paper/2504.09207/) (2025)
- [Table Search Using a Deep Contextualized Language Model](https://mlsanity.com/paper/2005.09207/) (2020)
- [Ad Hoc Table Retrieval using Semantic Similarity](https://mlsanity.com/paper/1802.06159/) (2018)
- [TARGET: Benchmarking Table Retrieval for Generative Tasks](https://mlsanity.com/paper/2505.11545/) (2025)
- [Auctus: A Dataset Search Engine for Data Augmentation](https://mlsanity.com/paper/2102.05716/) (2021)
- [NameGuess: Column Name Expansion for Tabular Data](https://mlsanity.com/paper/2310.13196/) (2023)
- [ExpandR: Teaching Dense Retrievers Beyond Queries with LLM Guidance](https://mlsanity.com/paper/2502.17057/) (2025)
- [Synthetic SQL Column Descriptions and Their Impact on Text-to-SQL Performance](https://mlsanity.com/paper/2408.04691/) (2024)
- [Automatic database description generation for Text-to-SQL](https://mlsanity.com/paper/2502.20657/) (2025)
- [Columbo: Expanding Abbreviated Column Names for Tabular Data Using Large Language Models](https://mlsanity.com/paper/2508.09403/) (2025)
- [Leveraging Schema Labels to Enhance Dataset Search](https://mlsanity.com/paper/2001.10112/) (2020)
- [Leveraging Retrieval Augmented Generative LLMs For Automated Metadata Description Generation to Enhance Data Catalogs](https://mlsanity.com/paper/2503.09003/) (2025)
- [WTR: A Test Collection for Web Table Retrieval](https://mlsanity.com/paper/2105.02354/) (2021)
- [Less Is More? When Dataset Context Hurts LLM-Generated Dataset Descriptions](https://mlsanity.com/paper/2606.02334/) (2026)
- [TACO: Task-Aware Column Description Generation Using LLMs](https://mlsanity.com/paper/2606.21685/) (2026)

## Abstract (arXiv)

Many table-centric NLP tasks such as NL2SQL first retrieve relevant tables from large collections using keyword search. Recent work uses LLMs to generate natural-language table descriptions to improve retrieval, but they are typically optimized for fluency rather than retrieval effectiveness. We present Polaris, a system that trains an LLM to generate table descriptions directly from retrieval feedback. Our key insight is that existing table retrieval benchmarks already contain the supervision needed for this task: given query-table relevance judgments, we generate multiple candidate descriptions for each table, rank them by their BM25 retrieval effectiveness, and use the resulting preference pairs to fine-tune the LLM with Direct Preference Optimization (DPO). Polaris further expands abbreviated table and column names before generation to reduce vocabulary mismatch. Extensive experiments show that Polaris outperforms the state-of-the-art AutoDDG solution, often by a significant margin. More broadly, our results demonstrate that retrieval benchmarks can be repurposed as supervision for training LLMs to generate retrieval-oriented metadata.