The Silent Cost of Conversation History in LLM Reasoning
Ali Şenol
Abstract
Large language models (LLMs) in multi-turn information systems accumulate their own earlier replies and a growing task state. We asked whether this conversation history lowers accuracy, and whether process-level quality metrics warn before it does. Each analysis was fixed before its outcomes were computed. In Study 1, on 48 sharded math tasks, retaining its replies lowered final success by 0.21 for GPT-4o-mini and 0.29 for Mistral-7B (Holm-adjusted p = 0.026 and 0.004); Llama-4 Scout showed the same direction, not significant. Study 2 introduces CDR-Stream, 100-turn state-tracking dialogues with an exact answer per turn and per-model difficulty calibration. With the full history, accuracy fell from about 0.6 to 0.1 within twenty turns at matched difficulty, whereas a solver-verified state summary kept it 0.52-0.59 higher (Holm-adjusted p less than 0.001). Consistency and logical coherence fell far less than accuracy, and stability rose. None of these metrics, a paired readiness measure or calibrated single-stream drift detectors gave robust, model-general early warning, although in Study 2 the repeat-based metrics were complete only for GPT-4o-mini. In the monitors we tested, a single 100-turn stream had power only for very large drops. Systems built on LLMs should keep task state outside the transcript and monitor across sessions.
Create a lesson
Related papers
Represented but misread: Post-training changes how language models read out what they represent
Jochen Laubrock
AutoMaster: Autonomous Post-Training by Making the Same Model a Better Teacher
Guoxin Chen, Shuang Sun, Fanzhe Meng et al.
Multimodal Error Detection in Chest X-ray Reports: A Large Language Model Reading Image Evidence
Yanda Cheng, Bo Wen, Yuhao Lu
MemFolder: Structure-Guided Memory Folding for Efficient Agent Memory Management
Xuyang Liu, Chenwen Lu, Junxi Wang et al.
Beam Model Technical Report
Reflection Team, Adam Abdulhamid, Osama Abuelsorour et al.
One Framework to Diagnose Them All: A Stratified Benchmark for Bengali LLM Hallucination
Asif Shahriyar Sushmit, Shefayat E Shams Adib, Abdullah Al Jubaer Gem et al.