mlsanityData to 14 Oct 2026

ClinTraceBench: Source-Verifiable Longitudinal Clinical Reasoning over EHR-Derived Dialogues

arXiv 2609.01111 ↗First posted 1 Sep 2026cs.CLcited by 1

On the map: LLM reasoning → Clinical and biomedical applications

also in 1 more
Abstract (arXiv)

Clinical LLM assistants must reason over multi-visit patient trajectories, yet whether the compact history representations used to scale them---retrieval, structured timelines, LLM summaries, agentic memory---preserve the longitudinal signal clinical reasoning needs has not been measured. We introduce ClinTraceBench: 385 MIMIC-IV-derived verified dialogues with event-ID provenance, a nine-task taxonomy (T1--T9), and L0--L4 deterministic + L5 human-audit validation (98.92% agreement). We evaluate eight history representation strategies---a no-context floor, textitlast-visit-only, textitfull-context, BGE-M3 textitdense-retrieval, two compression schemes, and two agentic-memory systems (textitMem0, textitA-Mem)---across four backbones (DeepSeek-V3, GPT-4o-mini, Haiku~4.5, Sonnet~4.6) on 6,271 questions: 32 cells, 200,672 predictions. Four findings: (SP4) a controlled T3 injection probe isolates compression-induced textitrelation loss---with the attribution sentence present textitbefore construction, textitMem0, textitA-Mem and textitllm-summary still recover only 0--5.3% of the injected positives; (SP1) compressed strategies pay an aggregation tax on multi-visit trends and cross-patient comparisons; (SP2) the blind-to-full gap spans +29.8~pp (GPT-4o-mini) to +62.7~pp (Haiku); (SP3) abstention scales non-monotonically with context length. On the Pareto frontier Haiku dominates Sonnet under textitfull-context (25.76 vs. 106.21), inverting the ``biggest backbone wins'' heuristic.

Key facts. Cited by 1 paper in the map (data through 14 Oct 2026). Placed in Clinical and biomedical applications, one of 17 directions in LLM reasoning. Builds on 9 of its 24 references that are in the map. First posted on arXiv 1 Sep 2026.

Builds on

9 of 24 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025Large Language Models with Temporal Reasoning for Longitudinal Clinical Summarization and Prediction30 Jan

Other references · 8

  1. 2022Language Models (Mostly) Know What They Knowcited by 1,856
  2. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
  3. 2025A-MEM: Agentic Memory for LLM Agentscited by 1,063
  4. 2025Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memorycited by 779
  5. 2024A Survey on the Memory Mechanism of Large Language Model based Agentscited by 634
  6. 2025Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactionscited by 254
  7. 2025MedHELM: Holistic Evaluation of Large Language Models for Medical Taskscited by 64
  8. 2024CliBench: A Multifaceted and Multigranular Evaluation of Large Language Models for Clinical Decision Makingcited by 8

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Personalized State-Transition-Aware Memory for Clinical Agentscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Beyond Prediction: Longitudinal Reasoning in EHR-Integrated Clinical AI ↗
  2. 2026Can LLMs Reason Over Long Horizons? An Empirical Evaluation of Context Strategies for Longitudinal Clinical Reasoning
  3. 2026ChatHealthAI: Aligning Electronic Health Record Representations with Large Language Models for Grounded Clinical Reasoning
  4. 2026GraphWalker: Patient Analogy Meets Information Gain for Clinical Reasoning with Large Language Models
  5. 2025TIMER: Temporal Instruction Modeling and Evaluation for Longitudinal Clinical Records
  6. 2026ClinicalBench: Stress-Testing Assertion-Aware Retrieval for Cross-Admission Clinical QA on MIMIC-IV
  7. 2026GLoC-EHR: Evidence-Cited Clinical Reasoning over Global Context and Local EHR Events
  8. 2026CliniCARE-Bench: Clinical Calibrated Audit of Medical Reasoning in EHR