Pith. sign in

Non-instructional Fine-tuning: Enabling Instruction-Following Capabilities in Pre-trained Language Models without Instruction-Following Data

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
abstract

Instruction fine-tuning is crucial for today's large language models (LLMs) to learn to follow instructions and align with human preferences. Conventionally, supervised data, including the instruction and the correct response, is required for instruction fine-tuning. To obtain such data, some researchers prompted well-trained models like GPT-4 to generate instructions and correct responses. In this paper, we propose a novel approach that uses the first half of a random text from OpenWebText as the instruction and GPT-3.5-turbo or GPT-4-turbo to complete the text as the response. Despite the data being "non-instructional", we found that pre-trained LLMs fine-tuned on this data can gain instruction-following capabilities. This observation is verified by fine-tuning several well-known pre-trained LLMs (e.g., LLaMA-2-7B, LLaMA-3-8B, LLaMA-3-70B, Mistral-7B-v0.1). The "non-instructional data" also improved some models that underwent supervised fine-tuning and human preference alignment. Our LLaMA-3-70B-Instruct fine-tuned through "non-instructional data" is comparable with LLaMA-3.1-70B-Instruct on the Arena Hard leaderboard. We analyzed the "non-instructional data" and ensured it is devoid of content related to instruction fine-tuning. Our findings will inspire further investigation into how to develop instruction-following capabilities without explicit instruction-related data.

fields

cs.CL 1

years

2025 1

verdicts

CONDITIONAL 1

representative citing papers

On the Effect of Instruction Tuning Loss on Generalization

cs.CL · 2025-07-10 · conditional · novelty 6.0

Weighted Instruction Tuning, with low-to-moderate prompt weight and moderate-to-high response weight, beats the standard response-only instruction tuning loss in most of the 75 (model, dataset, benchmark) settings tested.

citing papers explorer

Showing 1 of 1 citing paper.

  • On the Effect of Instruction Tuning Loss on Generalization cs.CL · 2025-07-10 · conditional · none · ref 53 · internal anchor

    Weighted Instruction Tuning, with low-to-moderate prompt weight and moderate-to-high response weight, beats the standard response-only instruction tuning loss in most of the 75 (model, dataset, benchmark) settings tested.