Lune

ACL Findings2026

Lost in the Prompt Order: Revealing the Limitations of Causal Attention in Language Models

Hyunjong Ok, Jaeho Lee

Abstract

Large language models exhibit surprising sensitivity to the structure of the prompt, but the mechanisms underlying this sensitivity remain poorly understood. In this work, we conduct an in-depth investigation on a striking case: in multiple-choice question answering, placing context before the questions and options (CQO) outperforms the reverse order (QOC) by over 14%p, consistently over a wide range of models and datasets. Through systematic architectural analysis, we identify causal attention as the core mechanism: in QOC prompts, the causal mask prevents option tokens from attending to context, creating an information bottleneck where context becomes invisible to options. Some Cantonese don't like chili, so some southerners don't like chili. A.Some Cantonese love chili B.Some people who like peppers are (...) C.All Cantonese are southerners D.Some Cantonese like neither peppers (...) Which of the following can guarantee the above argument? Some Cantonese don't like chili, so some southerners don't like chili. A.Some Cantonese love chili B.Some people who like peppers are (...) C.All Cantonese are southerners D.Some Cantonese like neither peppers (...) Which of the following can guarantee the above argument? Context Question Option QOC 14.7 % Drop!

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

Builds onCited byLeft to right by year; the more cited, the higher.

Builds on1

  1. Make Your LLM Fully Utilize the ContextShengnan An, Zexiong Ma, Zeqi Lin, Nanning Zheng et al.NeurIPS 2024 · 212 citations

Related papers

  1. COPO: Causal-Oriented Policy Optimization for Hallucinations of MLLMsPeizheng Guo, Jingyao Wang, Wenwen Qiang, Jiahuan Zhou et al.CVPR 2026 · 1 citation
  2. Rapid Selection and Ordering of In-Context Demonstrations via Prompt Embedding ClusteringKha Pham, Hung Le, Man Ngo, Truyen TranICLR 2025
  3. Artifacts or Abduction: How Do LLMs Answer Multiple-Choice Questions Without the Question?Nishant Balepur, Abhilasha Ravichander, Rachel RudingerACL 2024 · 5 citations
  4. Segment-Based Attention Masking for GPTsShahar Katz, Liran Ringel, Yaniv Romano, Lior WolfACL 2025
  5. Leveraging Large Language Models for Multiple Choice Question AnsweringJoshua Robinson, David WingateICLR 2023 · 40 citations
Dusk over the sea between two cliffs drawn in fine vertical lines