Beginner4 min read15 of 52

Instruct Model

A base model fine-tuned to follow instructions as a helpful assistant.

SPEC: Instruct Model

Definition

[Definition] An instruct model is a base model that has been further trained (via Supervised Fine-Tuning and/or RLHF) to understand and follow natural language instructions reliably. It responds to user requests rather than just completing text. Also called a "chat model" or "assistant model."

The Gap Between Base and Instruct

A base model trained to predict the next token knows a lot — but doesn't know it's supposed to be helpful. Fine-tuning on instruction-response pairs bridges this gap:

Base model sees: "Summarize this article: [article]..." → continues generating more article text Instruct model sees: "Summarize this article: [article]..." → generates a concise summary

How Instruct Models Are Created

Step 1: Supervised Fine-Tuning (SFT)

  • Collect a dataset of (instruction, ideal response) pairs
  • Fine-tune the base model to reproduce the ideal responses
  • Dataset is human-written or GPT-generated and human-verified
  • Result: model learns the instruction-following format

Step 2: RLHF / Preference Optimization (Optional but Standard)

  • Human raters compare model outputs and rank them
  • Train a reward model to predict human preferences
  • Use PPO (or DPO, ORPO) to optimize the model toward higher-reward outputs
  • Result: model becomes more helpful, honest, and harmless

Instruct Format (Chat Template)

Most instruct models use a structured prompt format:

<|system|> You are a helpful assistant. <|user|> What is photosynthesis? <|assistant|> Photosynthesis is the process by which plants...

Each model family has its own chat template (ChatML, Llama-3, Mistral, Gemma, etc.)

Key Capabilities Gained Through Instruction Tuning

CapabilityDescription
Instruction followingExecutes explicit user requests
Multi-turn conversationMaintains context across turns
FormattingFollows "respond in JSON", "use bullet points"
Role-playingAdopts personas when instructed
RefusalsDeclines harmful or out-of-scope requests
ConcisenessAnswers the question without rambling

Quality Dimensions of Instruct Models

DimensionDescription
HelpfulnessDoes it actually answer the question?
HarmlessnessDoes it avoid dangerous outputs?
HonestyDoes it express uncertainty appropriately?
Instruction adherenceDoes it follow all constraints in the prompt?
Format complianceDoes it match requested output format?

Popular Instruct Models

ModelBaseNotes
GPT-4oGPT-4 (base)OpenAI, closed
Claude 3.5 SonnetClaude (base)Anthropic, via API
LLaMA 3.1 InstructLLaMA 3.1Meta, open weights
Mistral InstructMistralOpen weights, efficient
Gemma 2 InstructGemma 2Google, open weights

Instruct Model vs. Chat Model

These terms are used interchangeably in practice. Some make a distinction:

  • Instruct model: single-turn task execution focus
  • Chat model: multi-turn conversational focus In practice, most modern models are both.

System Prompt Role

Instruct models accept a system prompt that shapes their behavior:

  • Persona definition: "You are an expert data scientist"
  • Behavioral constraints: "Always respond in French"
  • Safety rules: "Do not discuss competitors"
  • Tool/context injection: "You have access to the following tools: ..."

Tradeoffs of Instruction Tuning

[Key Insight] Instruction tuning makes models dramatically more useful for real-world tasks, but it comes with tradeoffs — the fine-tuning process can reduce raw capability compared to the base model, and safety-focused training may cause the model to refuse valid requests.

ProCon
Much more useful for usersCan reduce raw capability vs. base
Safer behaviorMay be overly cautious / refuse valid requests
Consistent formatMay lose some creative/unexpected completions
Better at following constraintsAlignment tax on some benchmarks

Related Concepts

  • Base Model, Fine-Tuning, RLHF, System Prompt, User Prompt, Alignment, SFT