LaRS: Latent Reasoning Skills for Chain-of-Thought Reasoning

Anonymous

Abstract: Chain-of-thought (CoT) prompting has emerged as a popular in-context learning (ICL) approach for large language models (LLMs), especially when tackling complex reasoning tasks. Traditional ICL approaches construct prompts using examples that contain questions similar to the input question. However, CoT prompting, which includes crucial intermediate reasoning steps (rationales) within its examples, necessitates selecting examples based on these rationales rather than the questions themselves. Existing methods require human experts or pre-trained LLMs to describe the skill, a high-level abstraction of rationales, to guide the selection. These methods, however, are often costly and difficult to scale. Instead, this paper introduces a new approach named Latent Reasoning Skills (LaRS) that employs unsupervised learning to create a latent space representation of rationales, with a latent variable called a reasoning skill. Concurrently, LaRS learns a reasoning policy to determine the required reasoning skill for a given question. Then the ICL examples are selected by aligning the reasoning skills between past examples and the question. Our approach is theoretically grounded and sample-efficient, eliminating the need for helper LLM inference or manual prompt design. Empirically, LaRS achieves performance comparable to SOTA rationale-based selection methods, saving thousands of LLM inferences and significantly reducing the time required to process the example bank.

Paper Type: long

Research Area: Machine Learning for NLP

Contribution Types: Approaches low compute settings-efficiency, Theory

Languages Studied: English

Preprint Status: There is a non-anonymous preprint (URL specified in the next question).

A1: yes

A1 Elaboration For Yes Or No: Section 7

A2: n/a

A3: yes

A3 Elaboration For Yes Or No: Abstract and Section 1

B: no

B1: n/a

B2: n/a

B3: n/a

B4: n/a

B5: n/a

B6: yes

B6 Elaboration For Yes Or No: Section 5

C: yes

C1: yes

C1 Elaboration For Yes Or No: Section 5

C2: yes

C2 Elaboration For Yes Or No: Section 5 and Appendix A

C3: yes

C3 Elaboration For Yes Or No: Section 5.2 and 5.3

C4: n/a

D: no

D1: n/a

D2: n/a

D3: n/a

D4: n/a

D5: n/a

E: no

E1: n/a