paper-with-me

Papers

RSAT: Structured Attribution Makes Small Language Models Faithful Table Reasoners

2026-04-30 · Jugal Gajjar, Kamalasankari Subramaniakuppusamy arxiv

When a language model answers a table question, users have no way to verify which cells informed which reasoning steps. We introduce RSAT, a method that trains small language models (SLMs, 1-8B) to produce step-by-step reasoning with cell-level citations grounded in table evidence. Phase 1 (SFT) teaches a structured JSON output format from verified reasoning traces. Phase 2 (GRPO) optimizes a composite reward centered on NLI-based faithfulness, alongside citation validity and parsimony. Across six models from two families-Qwen 2.5 (1.5B/3B/7B) and Llama 3 (1B/3B/8B)-RSAT improves faithfulness 3.7$\times$ over SFT alone (0.224$\rightarrow$0.826), with near-perfect citation validity (0.992). Post-hoc attribution collapses below 13% format success, confirming that attribution must be integrated into reasoning, not retrofitted. Ablations show the faithfulness reward is essential: removing it drops faithfulness from 0.97 to 0.03.

📄 PDF Abstract BibTeX arXiv:2605.00199

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

dIR -- Discrete Information Retrieval: Conversational Search over Unstructured (and Structured) Data with Large Language Models

2023-12-20 · Pablo M. Rodriguez Bertorello, Jean Rodmond Junior Laguerre

Data is stored in both structured and unstructured form. Querying both, to power natural language conversations, is a challenge. This paper introduces dIR, Discrete Information Retrieval, providing a unified interface to…

Conversational SearchInformation RetrievalLanguage ModelingLanguage Modelling+4

Abduct, Act, Predict: Scaffolding Causal Inference for Automated Failure Attribution in Multi-Agent Systems

2025-09-12 · Alva West, Yixuan Weng, Minjun Zhu, Zhen Lin 외 arxiv

Failure attribution in multi-agent systems -- pinpointing the exact step where a decisive error occurs -- is a critical yet unsolved challenge. Current methods treat this as a pattern recognition task over long conversat…

Causal Inference

Persistent and Conversational Multi-Method Explainability for Trustworthy Financial AI

2026-05-12 · Georgios Makridis, Georgios Fatouros, John Soldatos, George Katsis 외 arxiv

Financial institutions increasingly require AI explanations that are persistent, cross-validated across methods, and conversationally accessible to human decision-makers. We present an architecture for human-centered exp…

Semantic RetrievalSentiment Analysis

AttriBoT: A Bag of Tricks for Efficiently Approximating Leave-One-Out Context Attribution

2024-11-22 · Fengyuan Liu, Nikhil Kandpal, Colin Raffel

The influence of contextual input on the behavior of large language models (LLMs) has prompted the development of context attribution methods that aim to quantify each context span's effect on an LLM's generations. The l…

Topic Segmentation of Semi-Structured and Unstructured Conversational Datasets using Language Models

2023-10-26 · Reshmi Ghosh, Harjeet Singh Kajal, Sharanya Kamath, Dhuri Shrivastava 외

Breaking down a document or a conversation into multiple contiguous segments based on its semantic structure is an important and challenging problem in NLP, which can assist many downstream tasks. However, current works …

Segmentation