---
title: "LLMs Get Lost In Multi-Turn Conversation"
description: "LLMs Get Lost In Multi-Turn Conversation asks whether a model that solves a complete one-shot instruction remains dependable when the same requirements arrive over several turns. Read it to see how instruction sharding holds the intended task fixed, why best-case aptitude hides a 112% rise in unreliability, which recap and temperature interventions fall short, and where an idealized simulated user limits the result."
canonical_url: "https://fanout.sh/daily/2026-09-29-llms-lost-multi-turn-conversation"
md_url: "https://fanout.sh/daily/2026-09-29-llms-lost-multi-turn-conversation.md"
last_updated: "2026-09-29"
access: "public"
---

# LLMs Get Lost In Multi-Turn Conversation

LLMs Get Lost In Multi-Turn Conversation asks whether a model that solves a complete one-shot instruction remains dependable when the same requirements arrive over several turns. Read it to see how instruction sharding holds the intended task fixed, why best-case aptitude hides a 112% rise in unreliability, which recap and temperature interventions fall short, and where an idealized simulated user limits the result.

## Paper details

- Authors: Philippe Laban et al.

- Venue: ICLR 2026 Outstanding Paper

- Track: AI Research

- Difficulty: Easy

- Reading time: 36 min

- Original paper: https://arxiv.org/abs/2505.06120v1

## What you will learn

- How Full, Concat, and Sharded delivery separate instruction rewriting from multi-turn information timing.

- Why repeated simulations need both an aptitude measure and an unreliability range instead of one average score.

- How to read the 39% average degradation, partial mitigations, episodic translation control, and simulator limits without turning them into claims about natural user behavior.

## Continue exploring

[Compare the ReAct loop in Fanout](https://fanout.sh/daily/2026-08-10-react-reasoning-acting): Connect multi-turn requirement delivery to an explicit action-observation loop, where the model's own earlier outputs also become part of the next decision context.

---
This representation contains public Fanout content only. Protected Pro lessons, account data, billing, checkout, and pricing are not included.

Browse the public content map: https://fanout.sh/sitemap.md