SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2402.10646 · ACL · 2024

ABSINSTRUCT: Eliciting Abstraction Ability from LLMs through Explanation Tuning with Plausibility Estimation

Wei Fan, Xin Liu, Yangqiu Song, Hongming Zhang, Ginny Wong, Simon See, Tianqing Fang, Zhaowei Wang, Sehyun Choi, Qing Zong, Plausibility Estimator

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

Abstraction ability is crucial in human intelligence, which can also benefit various tasks in NLP study. Existing work shows that LLMs are deficient in abstract ability, and how to improve it remains unexplored. In this work, we design the framework ABSINSTRUCT to enhance LLMs' abstraction ability through instruction tuning. The framework builds instructions with in-depth explanations to assist LLMs in capturing the underlying rationale of abstraction. Meanwhile, we introduce a plausibility estimator to select instructions that are more consistent with the abstraction knowledge of LLMs to be aligned. Then, our framework combines abstraction instructions with general-purpose ones to build a hybrid dataset. Extensive experiments and analyses 1 demonstrate that our framework can considerably enhance LLMs' abstraction ability with strong generalization performance while maintaining their general instruction-following abilities. I cook a pie with apple -> fruit? Mary chews her food -> eats?

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2402.10646")
get_code_for_paper("2402.10646")
have("2402.10646")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2402.10646.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2402.10646.svg)](https://syntology.ai/paper/2402.10646)

Connect an agent — have() is free.