Instruction Diversity Drives Generalization To Unseen Tasks
- URL: http://arxiv.org/abs/2402.10891v1
- Date: Fri, 16 Feb 2024 18:47:21 GMT
- Title: Instruction Diversity Drives Generalization To Unseen Tasks
- Authors: Dylan Zhang, Justin Wang, Francois Charton
- Abstract summary: Generalization emerges once a diverse enough set of tasks is provided, even though very few examples are provided for each task.
Generalization emerges once a diverse enough set of tasks is provided, even though very few examples are provided for each task.
- Score: 1.9059113568275998
- License: http://creativecommons.org/licenses/by-sa/4.0/
- Abstract: Instruction tuning -- fine-tuning a large language model (LLM) on pairs of
instructions and desired outcomes -- is an approach that enables pre-trained
language models to perform real-world tasks and follow human instructions. Its
practical success depends on the model learning a broader set of instructions
than those it was trained on. Yet the factors that determine model
generalization to such \emph{unseen tasks} are not well understood. %To
understand the driving factors of generalization, In this paper, we experiment
with string rewrites, a symbolic task that serves as a building block for
Turing complete Markov algorithms while allowing experimental control of
"inputs" and "instructions". We investigate the trade-off between the number of
instructions the model is trained on and the number of training samples
provided for each instruction and observe that the diversity of the instruction
set determines generalization. Generalization emerges once a diverse enough set
of tasks is provided, even though very few examples are provided for each task.
Instruction diversity also ensures robustness with respect to non-uniform
distributions of instructions in the training set.
Related papers
- From Instance Training to Instruction Learning: Task Adapters Generation from Instructions [29.452006810725184]
This paper focuses on simulating human learning to address the shortcomings of instance training.
We introduce Task Adapters Generation from Instructions (TAGI), which automatically constructs the task-specific model.
We evaluate TAGI on the Super-Natural Instructions and P3 datasets.
arXiv Detail & Related papers (2024-06-18T08:14:28Z) - From Symbolic Tasks to Code Generation: Diversification Yields Better Task Performers [1.6958018695660049]
We show that a more diverse instruction set, extending beyond code-related tasks, improves the performance of code generation.
Our observations suggest that a more diverse semantic space for instruction-tuning sets greatly improves the model's ability to follow instructions and perform tasks.
arXiv Detail & Related papers (2024-05-30T07:54:07Z) - Chain-of-Instructions: Compositional Instruction Tuning on Large Language Models [15.444719480373001]
We propose a novel concept of compositional instructions called chain-of-instructions (CoI)
Unlike the conventional practice of solving single instruction tasks, our proposed method encourages a model to solve each subtask step by step until the final answer is reached.
CoI-tuning improves the model's ability to handle instructions composed of multiple subtasks as well as unseen composite tasks such as multilingual summarization.
arXiv Detail & Related papers (2024-02-18T10:10:40Z) - Context-dependent Instruction Tuning for Dialogue Response Generation [61.21790201307179]
Recent language models have achieved impressive performance in natural language computation tasks by incorporating instructions with task input during fine-tuning.
We introduce a context-based instruction fine-tuning framework for each multi-turn dialogue.
During the evaluation, the model generates instructions based on the previous context to self-guide the response.
arXiv Detail & Related papers (2023-11-13T01:25:30Z) - Specialist or Generalist? Instruction Tuning for Specific NLP Tasks [58.422495509760154]
We investigate whether incorporating broad-coverage generalist instruction tuning can contribute to building a specialist model.
Our experiments assess four target tasks with distinct coverage levels.
The effect is particularly pronounced when the amount of task-specific training data is limited.
arXiv Detail & Related papers (2023-10-23T19:46:48Z) - Instruction Position Matters in Sequence Generation with Large Language
Models [67.87516654892343]
Large language models (LLMs) are capable of performing conditional sequence generation tasks, such as translation or summarization.
We propose enhancing the instruction-following capability of LLMs by shifting the position of task instructions after the input sentences.
arXiv Detail & Related papers (2023-08-23T12:36:57Z) - Robustness of Learning from Task Instructions [15.462970803323563]
Traditional supervised learning mostly works on individual tasks and requires training on a large set of task-specific examples.
To build a system that can quickly and easily generalize to new tasks, task instructions have been adopted as an emerging trend of supervision.
This work investigates the system robustness when the instructions of new tasks are (i) manipulated, (ii) paraphrased, or (iii) from different levels of conciseness.
arXiv Detail & Related papers (2022-12-07T17:54:59Z) - Improving Task Generalization via Unified Schema Prompt [87.31158568180514]
Unified Prompt is a flexible and prompting method, which automatically customizes the learnable prompts for each task according to the task input schema.
It models the shared knowledge between tasks, while keeping the characteristics of different task schema.
The framework achieves strong zero-shot and few-shot performance on 16 unseen tasks downstream from 8 task types.
arXiv Detail & Related papers (2022-08-05T15:26:36Z) - Learning Action Conditions from Instructional Manuals for Instruction Understanding [48.52663250368341]
We propose a task dubbed action condition inference, and collecting a high-quality, human annotated dataset of preconditions and postconditions of actions in instructional manuals.
We propose a weakly supervised approach to automatically construct large-scale training instances from online instructional manuals, and curate a densely human-annotated and validated dataset to study how well the current NLP models can infer action-condition dependencies in instruction texts.
arXiv Detail & Related papers (2022-05-25T00:19:59Z) - Benchmarking Generalization via In-Context Instructions on 1,600+
Language Tasks [95.06087720086133]
Natural-Instructions v2 is a collection of 1,600+ diverse language tasks and their expert written instructions.
The benchmark covers 70+ distinct task types, such as tagging, in-filling, and rewriting.
This benchmark enables large-scale evaluation of cross-task generalization of the models.
arXiv Detail & Related papers (2022-04-16T03:12:30Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.