Skip to content
Preprint

A Unified Moral-Value Dataset for Instruction Tuning

Jul 2026 · 0 citations · 34 references
Computer Science

TL;DR

This work constructs a unified moral-value dataset that can be directly used for instruction tuning, built upon existing moral-value datasets by merging them into a unified corpus and converting them into an instruction-response format.

Abstract

Large language models (LLMs) have developed rapidly and become valuable tools in everyday life. However, how to align LLMs to a particular set of human values is still an open problem. Recent studies show that instruction tuning has strong potential for zero-shot tasks and may serve as an effective approach to addressing value alignment. Nevertheless, although many datasets for instruction tuning already exist, they are not specifically designed around moral scenarios and behaviors. We construct a unified moral-value dataset that can be directly used for instruction tuning. This dataset is built upon existing moral-value datasets by merging them into a unified corpus and converting them into an instruction-response format. We show that training on a mixed dataset combining general task datasets with our dataset preserves general-task performance, and we report preliminary observations on how the mixing ratio affects value-oriented task performance. Our work provides a moral-value dataset for instruction tuning and offers a useful resource for further alignment research. The dataset is available at https://huggingface.co/datasets/teohzzh/value-for-instruction-tuning.

View source

Similar papers

Aug 2026

STAR: instruction tuning for Arabic across tasks, datasets, and models

An in-depth evaluation of instruction tuning for Arabic NLP tasks using three prominent LLMs: LLaMA 3.1-8B, AceGPT-v2-8B, and Qwen3-8B shows that instruction tuning consistently improves performance across most tasks, with notable variations in effectiveness across different tasks and prompts.

Maged Saeed Al-shaibani, Zaid Alyafeai, Irfan Ahmad · 0 citations
Preprint Jul 2026

Understanding Axes of Difficulty For Long Context Tasks Via PredicateLongBench

PredicateLongBench is proposed, a benchmark that stress-tests long-context reasoning by asking models to identify the longest contiguous subsequence of words in a long input that satisfies given predicates/constraints drawn from a broader predicate class.

Siddhartha Jain, A. Velingker · 0 citations
Open access 2026

Asking language models how to represent data for fine-tuning

It is shown that format choice remains important even after fine-tuning; models learn more efficiently with specific formats rather than adapting to any format; this finding allows format selection to be done via inference alone, avoiding costly trial-and-error fine-tuning runs.

Usneek Singh, Ananya Singha, Abhijeet Awasthi et al. · 0 citations
Preprint Aug 2026

How Language Models Organize and Structure Moral Knowledge

This work trains six independent linear probes on open-weight language models, one per Moral Foundations Theory (MFT) category, and examines how the resulting directions relate to each other in representation space, finding the directions neither collapse into a single moral detector nor isolate from one another.

Orion Reblitz-Richardson · 0 citations
Preprint Aug 2026

Why Large Language Models Fail at Tabular Prediction

The results show that the LLM's capability dissolves with dimension in a way no noise-corrupted classical learner mimics - which explains why LLMs, so capable elsewhere, keep losing to fifty-year-old baselines on tables, while leaving the mechanism of the prediction as an open question.

M. Garnelo, Wojciech M. Czarnecki · 1 citation
Preprint Jul 2026

Selecting Open-Weight Language Models for Zero-Shot Intent Classification: A Systematic Evaluation of 41 Models

Intent classification is a core component of task-oriented dialogue systems, yet practitioners have limited systematic guidance for selecting deployable open-weight language models under compute, latency, and robustness constraints. We present a systematic zero-shot evaluation of 41 open-weight language models spanning 15 families and the 135M--9B parameter range across eight English single-label intent-classification datasets. A ninth dataset, ATIS, uses five labeled demonstrations and is reported as an auxiliary five-shot result. The evaluation includes standard benchmarks, a large-scale voice-assistant corpus, and production-derived e-commerce datasets. Beyond exact-match accuracy, we analyze confidence calibration, robustness to realistic input perturbations, statistical reliability of model rankings, deployment efficiency, and benchmark saturation. Our results show that instruction-tuned 3B models can outperform several evaluated 7B base models, that differences among leading models on MASSIVE are statistically indistinguishable under pairwise McNemar tests, and that widely used benchmarks such as SNIPS have become saturated and no longer meaningfully discriminate among current open-weight models. Instruction tuning's effect on confidence calibration is inconsistent rather than uniformly harmful. These findings provide practical guidance for selecting and evaluating open-weight language models for intent classification.

Parishruthi Ganesh, Gerry V. Dozier, Cheryl D. Seals · 0 citations