# Old Habits Die Hard: How Conversational History Geometrically Traps LLMs

arXiv: https://arxiv.org/abs/2603.03308 · HTML: https://mlsanity.com/paper/2603.03308/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 32 of its 44 references that are in the map.
- First posted on arXiv 8 Feb 2026.

## Builds on (most cited first, 30 of 32)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Qwen3 Technical Report](https://mlsanity.com/paper/2505.09388/) (2025)
- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehension](https://mlsanity.com/paper/1705.03551/) (2017)
- [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022)
- [Similarity of Neural Network Representations Revisited](https://mlsanity.com/paper/1905.00414/) (2019)
- [gpt-oss-120b & gpt-oss-20b Model Card](https://mlsanity.com/paper/2508.10925/) (2025)
- [Inference-Time Intervention: Eliciting Truthful Answers from a Language Model](https://mlsanity.com/paper/2306.03341/) (2023)
- [Qwen3 Embedding: Advancing Text Embedding and Reranking Through Foundation Models](https://mlsanity.com/paper/2506.05176/) (2025)
- [What Makes Good In-Context Examples for GPT-3?](https://mlsanity.com/paper/2101.06804/) (2021)
- [Towards Understanding Sycophancy in Language Models](https://mlsanity.com/paper/2310.13548/) (2023)
- [Refusal in Language Models Is Mediated by a Single Direction](https://mlsanity.com/paper/2406.11717/) (2024)
- [Large Language Models Cannot Self-Correct Reasoning Yet](https://mlsanity.com/paper/2310.01798/) (2023)
- [The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets](https://mlsanity.com/paper/2310.06824/) (2023)
- [The Internal State of an LLM Knows When It's Lying](https://mlsanity.com/paper/2304.13734/) (2023)
- [Safety Alignment Should Be Made More Than Just a Few Tokens Deep](https://mlsanity.com/paper/2406.05946/) (2024)
- [How Language Model Hallucinations Can Snowball](https://mlsanity.com/paper/2305.13534/) (2023)
- [Frontier Models are Capable of In-context Scheming](https://mlsanity.com/paper/2412.04984/) (2024)
- [SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal](https://mlsanity.com/paper/2406.14598/) (2024)
- [LoRA Fine-tuning Efficiently Undoes Safety Training in Llama 2-Chat 70B](https://mlsanity.com/paper/2310.20624/) (2023)
- [The Remarkable Robustness of LLMs: Stages of Inference?](https://mlsanity.com/paper/2406.19384/) (2024)
- [Similarity Analysis of Contextual Word Representation Models](https://mlsanity.com/paper/2005.01172/) (2020)
- [Large Language Models as Markov Chains](https://mlsanity.com/paper/2410.02724/) (2024)
- [The LLM Already Knows: Estimating LLM-Perceived Question Difficulty via Hidden Representations](https://mlsanity.com/paper/2509.12886/) (2025)
- [Belief Dynamics Reveal the Dual Nature of In-Context Learning and Activation Steering](https://mlsanity.com/paper/2511.00617/) (2025)
- [Sycophancy Hides Linearly in the Attention Heads](https://mlsanity.com/paper/2601.16644/) (2026)
- [Markov Chain of Thought for Efficient Mathematical Reasoning](https://mlsanity.com/paper/2410.17635/) (2024)
- [Agentic Uncertainty Quantification](https://mlsanity.com/paper/2601.15703/) (2026)
- [Layer by Layer: Uncovering Where Multi-Task Learning Happens in Instruction-Tuned Large Language Models](https://mlsanity.com/paper/2410.20008/) (2024)
- [Distinguishing Ignorance from Error in LLM Hallucinations](https://mlsanity.com/paper/2410.22071/) (2024)

## Built on by (most cited first, 6 of 6)

- [SWAY: A Counterfactual Computational Linguistic Approach to Measuring and Mitigating Sycophancy](https://mlsanity.com/paper/2604.02423/) (2026)
- [Most LLM Conformity Needs No Speaker: Measuring the Speaker-Free Floor in Peer-Pressure Benchmarks](https://mlsanity.com/paper/2607.05545/) (2026)
- [Recalling Too Well: Sycophancy Evaluation and Mitigation in Memory-Augmented Models](https://mlsanity.com/paper/2606.10949/) (2026)
- [Topics as Proxies for Sociodemographics: How Conversational Context Affects LLM Answers](https://mlsanity.com/paper/2606.02776/) (2026)
- [Decomposing how prompting steers behavior](https://mlsanity.com/paper/2606.03093/) (2026)
- [AMEL: Accumulated Message Effects on LLM Judgments](https://mlsanity.com/paper/2605.22714/) (2026)

## Abstract (arXiv)

How does the conversational past of large language models (LLMs) influence their future performance? Recent work suggests that LLMs are affected by their conversational history in unexpected ways. For instance, hallucinations in prior interactions may influence subsequent model responses. In this work, we introduce History-Echoes, a framework that investigates how conversational history biases subsequent generations. The framework explores this bias from two perspectives: probabilistically, we model conversations as Markov chains to quantify state consistency; geometrically, we measure the consistency of consecutive hidden representations. Across three model families and six datasets spanning diverse phenomena, our analysis reveals a strong correlation between the two perspectives. By bridging these perspectives, we demonstrate that behavioral persistence manifests as a geometric trap, where gaps in the latent space confine the model's trajectory. Code available at https://github.com/technion-cs-nlp/OldHabitsDieHard.