Timezone: »

Post Hoc Explanations of Language Models Can Improve Language Models
Satyapriya Krishna · Jiaqi Ma · Dylan Slack · Asma Ghandeharioun · Sameer Singh · Himabindu Lakkaraju

Tue Dec 12 03:15 PM -- 05:15 PM (PST) @ Great Hall & Hall B1+B2 #1424

Large Language Models (LLMs) have demonstrated remarkable capabilities in performing complex tasks. Moreover, recent research has shown that incorporating human-annotated rationales (e.g., Chain-of-Thought prompting) during in-context learning can significantly enhance the performance of these models, particularly on tasks that require reasoning capabilities. However, incorporating such rationales poses challenges in terms of scalability as this requires a high degree of human involvement. In this work, we present a novel framework, Amplifying Model Performance by Leveraging In-Context Learning with Post Hoc Explanations (AMPLIFY), which addresses the aforementioned challenges by automating the process of rationale generation. To this end, we leverage post hoc explanation methods which output attribution scores (explanations) capturing the influence of each of the input features on model predictions. More specifically, we construct automated natural language rationales that embed insights from post hoc explanations to provide corrective signals to LLMs. Extensive experimentation with real-world datasets demonstrates that our framework, AMPLIFY, leads to prediction accuracy improvements of about 10-25% over a wide range of tasks, including those where prior approaches which rely on human-annotated rationales such as Chain-of-Thought prompting fall short. Our work makes one of the first attempts at highlighting the potential of post hoc explanations as valuable tools for enhancing the effectiveness of LLMs. Furthermore, we conduct additional empirical analyses and ablation studies to demonstrate the impact of each of the components of AMPLIFY, which, in turn, lead to critical insights for refining in context learning.

Author Information

Satyapriya Krishna (Harvard University)
Jiaqi Ma (University of Illinois Urbana-Champaign)
Dylan Slack (Scale AI)
Asma Ghandeharioun (Google Research)
Sameer Singh (University of California, Irvine)

Sameer Singh is an Assistant Professor at UC Irvine working on robustness and interpretability of machine learning. Sameer has presented tutorials and invited workshop talks at EMNLP, Neurips, NAACL, WSDM, ICLR, ACL, and AAAI, and received paper awards at KDD 2016, ACL 2018, EMNLP 2019, AKBC 2020, and ACL 2020. Website: http://sameersingh.org/

Himabindu Lakkaraju (Harvard)

More from the Same Authors