LLM 기반 Agent의 Memory 메커니즘 종합 Survey

# Reflexion 스타일 cross-trial 메모리 패턴 예시

from openai import OpenAI

client = OpenAI()

# 실패 경험을 언어로 저장하는 메모리 저장소
cross_trial_memory = []

def reflect_on_failure(task, failed_trajectory, error_feedback):
    """실패 경험을 분석해 교훈을 추출"""
    prompt = f"""
    Task: {task}
    What I tried: {failed_trajectory}
    What went wrong: {error_feedback}
    
    In 2-3 sentences, what should I do differently next time?
    """
    response = client.chat.completions.create(
        model="gpt-4",
        messages=[{"role": "user", "content": prompt}]
    )
    return response.choices[0].message.content

def run_agent_with_memory(task, max_trials=3):
    for trial in range(max_trials):
        # 과거 실패 경험을 프롬프트에 포함
        memory_context = ""
        if cross_trial_memory:
            memory_context = "Past experiences:\n" + "\n".join(
                [f"- {m}" for m in cross_trial_memory[-3:]]  # 최근 3개만
            )
        
        prompt = f"""
        {memory_context}
        
        Now solve this task: {task}
        """
        
        response = client.chat.completions.create(
            model="gpt-4",
            messages=[{"role": "user", "content": prompt}]
        )
        
        result = response.choices[0].message.content
        
        # 실패 시 reflection 후 메모리에 저장
        success = evaluate_result(result)  # 실제 평가 로직
        if not success:
            lesson = reflect_on_failure(task, result, "Task failed")
            cross_trial_memory.append(lesson)
            print(f"Trial {trial+1} failed. Learned: {lesson}")
        else:
            return result
    
    return "Max trials reached"

# Retrieved Memory 패턴 (FAISS 활용)
import faiss
import numpy as np

class AgentMemory:
    def __init__(self, embedding_dim=1536):
        self.index = faiss.IndexFlatL2(embedding_dim)
        self.memories = []
    
    def write(self, text, embedding):
        """메모리 저장"""
        self.index.add(np.array([embedding]))
        self.memories.append(text)
    
    def read(self, query_embedding, top_k=3):
        """관련 메모리 검색"""
        distances, indices = self.index.search(
            np.array([query_embedding]), top_k
        )
        return [self.memories[i] for i in indices[0] if i < len(self.memories)]