# Finding Generalizable Evidence by Learning to Convince Q&A Models

arXiv: https://arxiv.org/abs/1909.05863 · HTML: https://mlsanity.com/paper/1909.05863/

## Key facts

- Cited by 25 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 12 Sep 2019.

## Built on by (most cited first, 25 of 25)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [Fine-Tuning Language Models from Human Preferences](https://mlsanity.com/paper/1909.08593/) (2019)
- [Red Teaming Language Models with Language Models](https://mlsanity.com/paper/2202.03286/) (2022)
- [Improving alignment of dialogue agents via targeted human judgements](https://mlsanity.com/paper/2209.14375/) (2022)
- [Self-critiquing models for assisting human evaluators](https://mlsanity.com/paper/2206.05802/) (2022)
- [Teaching language models to support answers with verified quotes](https://mlsanity.com/paper/2203.11147/) (2022)
- [Recursively Summarizing Books with Human Feedback](https://mlsanity.com/paper/2109.10862/) (2021)
- [Explaining Answers with Entailment Trees](https://mlsanity.com/paper/2104.08661/) (2021)
- [Chain of Hindsight Aligns Language Models with Feedback](https://mlsanity.com/paper/2302.02676/) (2023)
- [State-of-the-art generalisation research in NLP: A taxonomy and review](https://mlsanity.com/paper/2210.03050/) (2022)
- [e-CARE: a New Dataset for Exploring Explainable Causal Reasoning](https://mlsanity.com/paper/2205.05849/) (2022)
- [Rissanen Data Analysis: Examining Dataset Characteristics via Description Length](https://mlsanity.com/paper/2103.03872/) (2021)
- [Sample Efficient Preference Alignment in LLMs via Active Exploration](https://mlsanity.com/paper/2312.00267/) (2023)
- [Robustifying Multi-hop QA through Pseudo-Evidentiality Training](https://mlsanity.com/paper/2107.03242/) (2021)
- [ChatGPT vs State-of-the-Art Models: A Benchmarking Study in Keyphrase Generation Task](https://mlsanity.com/paper/2304.14177/) (2023)
- [JASMINE: Arabic GPT Models for Few-Shot Learning](https://mlsanity.com/paper/2212.10755/) (2022)
- [What's in your Head? Emergent Behaviour in Multi-Task Transformer Models](https://mlsanity.com/paper/2104.06129/) (2021)
- [Tournament of Prompts: Evolving LLM Instructions Through Structured Debates and Elo Ratings](https://mlsanity.com/paper/2506.00178/) (2025)
- [Kernelized Offline Contextual Dueling Bandits](https://mlsanity.com/paper/2307.11288/) (2023)
- [Towards Trustworthy Multimodal Moderation via Policy-Aligned Reasoning and Hierarchical Labeling](https://mlsanity.com/paper/2508.03296/) (2025)
- [Anchored Alignment for Self-Explanations Enhancement](https://mlsanity.com/paper/2410.13216/) (2024)
- [Reward Design for Justifiable Sequential Decision-Making](https://mlsanity.com/paper/2402.15826/) (2024)
- [Self-supervised Preference Optimization: Enhance Your Language Model with Preference Degree Awareness](https://mlsanity.com/paper/2409.17791/) (2024)

## Abstract (arXiv)

We propose a system that finds the strongest supporting evidence for a given answer to a question, using passage-based question-answering (QA) as a testbed. We train evidence agents to select the passage sentences that most convince a pretrained QA model of a given answer, if the QA model received those sentences instead of the full passage. Rather than finding evidence that convinces one model alone, we find that agents select evidence that generalizes; agent-chosen evidence increases the plausibility of the supported answer, as judged by other QA models and humans. Given its general nature, this approach improves QA in a robust manner: using agent-selected evidence (i) humans can correctly answer questions with only ~20% of the full passage and (ii) QA models can generalize to longer passages and harder questions.