LLMs Get Lost In Multi-Turn Conversation · HackerPapers discussions

LLMs Get Lost In Multi-Turn Conversation — mentioned by lkbm

Mon, 08 Dec 2025 16:48:20 GMT

There a recent(ish: May 2025) paper about how drip-feeding information is worse than restarting with a revised prompt once you realize details are missing.[0] [0] https://arxiv.org/abs/2505.06120

Paper and discussion history. A mention is not an endorsement.

LLMs Get Lost In Multi-Turn Conversation — mentioned by mrcoders

Wed, 02 Jul 2025 17:59:37 GMT

Looking at this thread, excited to see the community articulating what many of us have been experiencing. The shift from "prompt engineering" to "context engineering" really captures what's happening. The technical stuff (RAG, vector databases) is getting commoditized. But there's this foundational knowledge organization layer that's becoming critical. Most companies have context scattered across wikis, Slack, Google Docs. You can build sophisticated retrieval systems, but if you're feeding them fragmented information, you're missing huge optimization opportunities. There's research backing…

Paper and discussion history. A mention is not an endorsement.

LLMs Get Lost In Multi-Turn Conversation — mentioned by hbbio

Fri, 16 May 2025 03:17:19 GMT

That's why in practice you need more than this simple loop! Pretty much WIP, but I am experimenting with simple sequence-based workflows that are designed to frequently reset the conversation [2] This goes well with Microsoft paper "LLMs Get Lost In Multi-Turn Conversation " that was published Friday [1]. - [1]: https://arxiv.org/abs/2505.06120 - [2]: https://github.com/hbbio/nanoagent/blob/main/src/workflow.ts

Paper and discussion history. A mention is not an endorsement.

LLMs Get Lost In Multi-Turn Conversation — HN submission

Thu, 15 May 2025 02:28:42 GMT

LLMs get lost in multi-turn conversation

374 points; 259 discussion replies on HN. Paper and source history. A submission is not an endorsement.