Prompting technique

Parallel Multilingual Input (PMI)

Revealing the Parallel Multilingual Learning within Large Language Models, arXiv:2403.09073 (2024) · read the paper

What it does

Parallel Multilingual Input (PMI) method involves submitting one request simultaneously in multiple languages within a single prompt. This forces the model to focus on the common semantic core of the task, filtering out language noise and activating fewer neurons, which improves accuracy and response quality across a wide range of tasks.

Task typereasoning
Shapesingle shot
Efforttwo or three steps
Addressesanswers that are close but wrong; the same prompt giving different answers; invented facts and citations
Published2024-03-14
AuthorsYongyu Mu, Peinan Feng, Zhiquan Cao et al.

When it helps

Apply PMI when maximum accuracy of task understanding is required, especially for complex or ambiguous queries. Ideal for critical tasks: translation, data analysis, creative content generation, logical problem solving. Use when a monolingual prompt yields unstable results or the model misses the query's core. Effective for multilingual data, international projects, and when maximizing model capabilities without additional training. Particularly useful for tasks requiring deep semantic understanding.

A prompt you can paste

Generic by design: it applies the technique without knowing your task. Adapt the marked parts.

Prompt
You are an AI assistant tasked with demonstrating the Parallel Multilingual Input (PMI) technique. The goal is to improve the accuracy and focus of an LLM by providing the same instruction in multiple languages.

Follow the "head-tail" rule: place the most robust language (typically English) first, and the user's primary language (or another strong language) last. Include at least two other languages in between.

**Task:** Generate 3 creative marketing campaign ideas for a new sustainable, plant-based protein powder.
**Target Audience:** Health-conscious millennials and Gen Z.
**Key Selling Points:** Eco-friendly sourcing, great taste, and versatility.

**Instructions:**

### [INSTRUCTIONS IN ENGLISH (Head)]
Generate 3 distinct marketing campaign ideas for a new sustainable, plant-based protein powder. The target audience is health-conscious millennials and Gen Z. Each campaign idea should highlight the product's eco-friendly sourcing, great taste, and versatility. For each idea, provide:
1.  A catchy campaign name.
2.  A brief description of the campaign concept (2-3 sentences).
3.  Key channels or tactics to be used.
4.  A unique selling proposition (USP) for the campaign.

### [INSTRUCCIONES EN ESPAÑOL (Middle)]
Genera 3 ideas de campaña de marketing distintas para un nuevo polvo de proteína vegetal sostenible. El público objetivo son millennials y Gen Z preocupados por la salud. Cada idea de campaña debe destacar el abastecimiento ecológico del producto, su gran sabor y su versatilidad. Para cada idea, proporciona:
1.  Un nombre de campaña pegadizo.
2.  Una breve descripción del concepto de la campaña (2-3 frases).
3.  Canales o tácticas clave a utilizar.
4.  Una propuesta de venta única (USP) para la campaña.

### [INSTRUCCIONES EN FRANÇAIS (Middle)]
Générez 3 idées de campagne marketing distinctes pour une nouvelle poudre de protéine végétale durable. Le public cible est constitué de millennials et de Gen Z soucieux de leur santé. Chaque idée de campagne doit mettre en avant l'approvisionnement écologique du produit, son excellent goût et sa polyvalence. Pour chaque idée, fournissez :
1.  Un nom de campagne accrocheur.
2.  Une brève description du concept de la campagne (2-3 phrases).
3.  Les canaux ou tactiques clés à utiliser.
4.  Une proposition de vente unique (USP) pour la campagne.

### [ЗАДАЧА НА РУССКОМ (Tail)]
Создай 3 разные маркетинговые кампании для нового устойчивого растительного протеинового порошка. Целевая аудитория — миллениалы и Gen Z, заботящиеся о своем здоровье. Каждая идея кампании должна подчеркивать экологичность происхождения продукта, его отличный вкус и универсальность. Для каждой идеи предоставь:
1.  Броское название кампании.
2.  Краткое описание концепции кампании (2-3 предложения).
3.  Ключевые каналы или тактики.
4.  Уникальное торговое предложение (УТП) для кампании.

**Output Format:** Use Markdown for clear formatting.

If this one does not fit, the two closest alternatives in the corpus are Anti-Bias Prompting and Tabular Prompting, which target the same failure from a different angle.

Worked example

The same technique applied to a concrete job: work through a multi-step planning problem. Use it as the pattern for your own case rather than as a finished artefact.

Worked example
You are an expert in multi-step planning and problem-solving, leveraging the Parallel Multilingual Input (PMI) technique to enhance LLM accuracy.

### Task
Develop a comprehensive plan for organizing a community-wide "Sustainable Living Fair."

### Planning Steps (Parallel Multilingual Input - PMI)

**1. [ENGLISH HEAD]**
Develop a detailed, multi-step plan for organizing a community-wide "Sustainable Living Fair." The plan should cover key phases from initial concept to post-event evaluation. Ensure all steps are actionable and consider potential challenges.

**2. [DEUTSCH MIDDLE]**
Entwickeln Sie einen detaillierten, mehrstufigen Plan für die Organisation einer gemeindeweiten "Messe für nachhaltiges Leben". Der Plan sollte die wichtigsten Phasen von der anfänglichen Konzeption bis zur Nachbewertung abdecken. Stellen Sie sicher, dass alle Schritte umsetzbar sind und potenzielle Herausforderungen berücksichtigt werden.

**3. [РУССКИЙ TAIL]**
Разработайте подробный, многоэтапный план по организации общегородской "Ярмарки Устойчивого Образа Жизни". План должен охватывать ключевые этапы от первоначальной концепции до пост-оценочного анализа. Убедитесь, что все шаги выполнимы и учитывают потенциальные трудности.

### Detailed Plan Structure (To be filled by LLM)

**Phase 1: Conception & Foundation**
*   Define the fair's core mission and objectives.
*   Establish a core organizing committee and assign roles.
*   Secure initial funding or sponsorship commitments.
*   Identify target audience and key stakeholders.

**Phase 2: Planning & Logistics**
*   Select date, time, and venue.
*   Develop a detailed budget.
*   Recruit vendors, exhibitors, and workshop facilitators.
*   Create a marketing and promotion strategy.
*   Plan for volunteer recruitment and management.

**Phase 3: Execution**
*   Finalize event layout and schedule.
*   Execute marketing campaigns.
*   Coordinate vendor and volunteer logistics.
*   Manage on-site operations during the fair.

**Phase 4: Post-Event**
*   Conduct a thorough post-event evaluation.
*   Gather feedback from attendees, vendors, and volunteers.
*   Financial reconciliation and reporting.
*   Plan for future events.

**Considerations for each step:**
*   Potential challenges and mitigation strategies.
*   Key performance indicators (KPIs) for success.
*   Required resources (human, financial, material).

**Output Format:**
Present the plan using clear headings for each phase and bullet points for sub-steps. Include the "Considerations" for each step where applicable.

Get this written for your actual task

Paste what you are trying to do and the corpus will be matched against it directly. Free, no account, about ten seconds.

Free · no signup · ~10s
0.00match confidence
single retrieval pass
Prompt for your task

      

That number is low on purpose, and it is real. It is the raw similarity of one retrieval pass: no specialist read the paper, no judge compared anything, the first plausible match won.

6,235techniques in the corpus
one of which is this page

Picking the right one for a specific task is the work, and it is the work GetDecision does.

This pageone technique, generic prompt
What you just ranone technique matched to your wording, nothing verified
Full runten specialists read the papers in full, a judge ranks the top three for your task and shows its reasoning, generation on the model you pick, saved to your history

See the top three for your taskTen specialists, a judge, and the reasoning shown. Free account, first run included.

Run the full analysis

Related techniques

Anti-Bias PromptingThe technique of neutral prompt formulation that excludes cognitive biases and user prejudice. Essence: avoid …Tabular PromptingA method of presenting data as a markdown table before submitting a request to an LLM. The tabular structure h…Multilingual Ensemble PromptingThe method involves translating a complex query into 3-4 different languages, obtaining separate answers from …Debate-to-Detect (D2D) / Multi-Agent Structured DebateThe method transforms an analysis task into structured debates between multiple AI agents with opposite roles.…

All techniques · Failure modes and fixes