# Task Competence Is Not Instruction Following: Evaluating Instruction-Conflicting Behavior in Small Language Models

arXiv: https://arxiv.org/abs/2607.19608 · HTML: https://mlsanity.com/paper/2607.19608/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 23 of its 31 references that are in the map.
- First posted on arXiv 21 Jul 2026.

## Builds on (most cited first, 23 of 23)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [Finetuned Language Models Are Zero-Shot Learners](https://mlsanity.com/paper/2109.01652/) (2021)
- [Can a Suit of Armor Conduct Electricity? A New Dataset for Open Book Question Answering](https://mlsanity.com/paper/1809.02789/) (2018)
- [Multitask Prompted Training Enables Zero-Shot Task Generalization](https://mlsanity.com/paper/2110.08207/) (2021)
- [Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference](https://mlsanity.com/paper/1902.01007/) (2019)
- [Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Tasks](https://mlsanity.com/paper/2204.07705/) (2022)
- [RACE: Large-scale ReAding Comprehension Dataset From Examinations](https://mlsanity.com/paper/1704.04683/) (2017)
- [The Flan Collection: Designing Data and Methods for Effective Instruction Tuning](https://mlsanity.com/paper/2301.13688/) (2023)
- [FinBERT: Financial Sentiment Analysis with Pre-trained Language Models](https://mlsanity.com/paper/1908.10063/) (2019)
- [Solving General Arithmetic Word Problems](https://mlsanity.com/paper/1608.01413/) (2016)
- [Do Prompt-Based Models Really Understand the Meaning of their Prompts?](https://mlsanity.com/paper/2109.01247/) (2021)
- [A Diverse Corpus for Evaluating and Developing English Math Word Problem Solvers](https://mlsanity.com/paper/2106.15772/) (2021)
- [Evaluating Large Language Models at Evaluating Instruction Following](https://mlsanity.com/paper/2310.07641/) (2023)
- [Inverse Scaling: When Bigger Isn't Better](https://mlsanity.com/paper/2306.09479/) (2023)
- [Leveraging Large Language Models for Multiple Choice Question Answering](https://mlsanity.com/paper/2210.12353/) (2022)
- [A Survey of Deep Learning for Mathematical Reasoning](https://mlsanity.com/paper/2212.10535/) (2022)
- [Qwen-Scope: Turning Sparse Features into Development Tools for Large Language Models](https://mlsanity.com/paper/2605.11887/) (2026)
- [The SIFo Benchmark: Investigating the Sequential Instruction Following Ability of Large Language Models](https://mlsanity.com/paper/2406.19999/) (2024)
- [Wait, that's not an option: LLMs Robustness with Incorrect Multiple-Choice Options](https://mlsanity.com/paper/2409.00113/) (2024)
- [Deep Learning-based Sentiment Classification: A Comparative Survey](https://mlsanity.com/paper/2312.17253/) (2023)

## Abstract (arXiv)

Instruction tuning is meant to make language models follow user requests, yet it is unclear whether small models comply when an instruction conflicts with their usual task behavior. We study this across three tasks - multiple-choice question answering (MCQA), sentiment classification, and mathematical question answering - by pairing a standard instruction with a conflicting non-standard one (select an incorrect option, output the opposite sentiment, or return twice the answer). This cross-task design allows us to test whether resistance to conflicting instructions is tied to specific task characteristics or reflects a broader behavioral tendency. As all predictions are scored against the original ground truth, a model that ignores the non-standard instruction still appears accurate. Using standard accuracy, non-standard accuracy, and an Instruction-Following Failure Rate (IFFR), we evaluate instruction-tuned Qwen models across sizes. Both standard accuracy and instruction following generally improve with scale, although the pattern is not consistent across all tasks and datasets. Small models stay competent yet routinely ignore the non-standard instruction, while larger models show a clear gap between the two settings. These findings suggest that gains in task capability do not automatically provide reliable control over model behavior. Task competence and instruction following are therefore distinct abilities, and reporting only standard accuracy hides instruction-following failures.