From Exposure to Internalization: Dual-Stream Calibration for In-context Clinical Reasoning

Chuang Zhao; Hongke Zhao; Xiaofang Zhou; Xiaomeng Li

q-bio.QMApr 7, 2026

From Exposure to Internalization: Dual-Stream Calibration for In-context Clinical Reasoning

Authors: Chuang Zhao, Hongke Zhao, Xiaofang Zhou, Xiaomeng Li

Organizations: Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology, Hong Kong, SAR, China · College of Management and Economics, Laboratory of Computation and Analytics of Complex Management Systems (CACMS), Tianjin University, Tianjin 30072, China · Department of Computer Science and Engineering, The Hong Kong University of Science and Technology, Hong Kong, SAR, China

Abstract

Contextual clinical reasoning demands robust inference grounded in complex, heterogeneous clinical records. While state-of-the-art fine-tuning, in-context learning (ICL), and retrieval-augmented generation (RAG) enable knowledge exposure, they often fall short of genuine contextual internalization: dynamically adjusting a model's internal representations to the subtle nuances of individual cases at inference time. To address this, we propose Dual-Stream Calibration (DSC), a test-time training framework that transcends superficial knowledge exposure to achieve deep internalization during inference. DSC facilitates input internalization by synergistically aligning two calibration streams. Unlike passive context exposure, the Semantic Calibration Stream enforces a deliberative reflection on core evidence, internalizing semantic anchors by minimizing entropy to stabilize generative trajectories. Simultaneously, the Structural Calibration Stream assimilates latent inferential dependencies through an iterative meta-learning objective. By training on specialized support sets at test-time, this stream enables the model to bridge the gap between external evidence and internal logic, synthesizing fragmented data into a coherent response. Our approach shifts the reasoning paradigm from passive attention-based matching to an active refinement of the latent inferential space. Validated against thirteen clinical datasets, DSC demonstrates superiority across three distinct task paradigms, consistently outstripping state-of-the-art baselines ranging from training-dependent models to test-time learning frameworks.

Figures & tables

Panel of Figure 1 — Fig. 1: Motivations. Fig. 1(a) presents the ROUGE-L scores of various models evaluated on auxiliary QA pairs per sample. These QA pairs are generated from the eLife dataset’s context using DeepSeek-V3 [ 6 ] and subsequently provided as input for model evaluation. Qwen [ 7 ] and Lingshu [ 8 ] denote different LLM backbones. Fig. 1(b) depicts the predictive entropy fluctuations during the generative process. Fig. 1(c) evaluates model resilience (ROUGE-L) to contextual order on the eLife dataset, where the order of in-context demonstrations is randomly perturbed.

Fig. 2: Key difference between ICL, RAG, TTL-based baselines, and the proposed DSC. Our core innovation introduces a dual-calibration framework that enforces semantic understanding and structure robustness, transforming passive information retrieval into active contextual internalization.

Notations	Descriptions
$\mathcal{M}(\theta)$	Frozen Large Language Model
$\mathcal{C},\mathcal{Q},\mathcal{A}$	Retrieved Context, Input Query, Answer
$\mathbf{X}$	Total Input
$\mathbf{H}_{\mathcal{C}},\mathbf{H}_{\mathcal{Q}}$	Token Embeddings of $\mathcal{C}$ and $\mathcal{Q}$
$\boldsymbol{\delta}_{\text{sem}},\boldsymbol{\delta}_{\text{str}}$	Semantic/Structure Calibration Vectors (Trainable)
$m,n,k$	Sequence Lengths of $\mathcal{C}$ , $\mathcal{Q}$ , and $\mathcal{A}$

TABLE I: Comprehensive Mathematical Notations.

Fig. 3: Overview of our DSC framework. The overall architecture of DSC is presented in (a), which outlines the comprehensive test-time training pipeline, while (b) provides a granular decomposition of our core dual-stream mechanism. DSC leverages two parallel streams to simultaneously refine the input from both semantic and structural dimensions. The Semantic Calibration Stream implements dynamic entropy detection and elimination. This process first localizes high-uncertainty tokens via long-short window entropy analysis and subsequently optimizes the model using dual objectives: an entropy loss to resolve ambiguity and a high-fidelity loss to preserve critical information, resulting in a semantically certain input representation. The Structure Calibration Stream runs in parallel to reconstruct the vital inferential links bridging the context and the clinical query. It employs an iterative, alternating framework where specialized support sets and meta-queries are dynamically synthesized for meta-optimization. Finally, (c) illustrates the test-time inference phase, where the model executes the final predictions using the parameters optimized specifically for the given query. $\theta^{*}$ denotes the LLM parameters integrated with $\delta$ .

Methods	MedQA	PubMedQA	MedMCQA	MedBullets	MMLU	MMLU-Pro	MedExQA
Qwen2.5-7B [ 7 ]	0.160	0.160	0.240	0.045	0.127	0.260	0.090
Deepseek-R1-7B [ 6 ]	0.180	0.230	0.260	0.134	0.260	0.100	0.210
Lingshu-7B [ 8 ]	0.190	0.170	0.260	0.157	0.273	0.210	0.210
SFT [ 9 ]	0.210	0.180	0.080	0.184	0.260	0.180	0.160
GRPO [ 11 ]	0.230	0.240	0.230	0.134	0.150	0.180	0.180
ICL [ 17 ]	0.230	0.190	0.200	0.168	0.219	0.250	0.150

TABLE II: Performance comparison: Examination QA (HARD VERSION [ 33 ] ). The pipeline integrates Qwen2.5-7B [ 7 ] (LLM backbone), with further variations explored in Section V-B .

Methods	Cochrane			eLife			PLOS
Metrics	ROUGE-1	ROUGE-L	SARI	ROUGE-1	ROUGE-L	SARI	ROUGE-1	ROUGE-L	SARI
Qwen2.5-7B [ 7 ]	0.397	0.372	0.384	0.350	0.322	0.433	0.370	0.348	0.394
Deepseek-R1-7B [ 6 ]	0.400	0.369	0.383	0.368	0.349	0.437	0.410	0.372	0.367
Lingshu-7B [ 8 ]	0.385	0.357	0.383	0.312	0.291	0.428	0.357	0.329	0.384
SFT [ 9 ]	0.435	0.400	0.382	0.423	0.409	0.405	0.415	0.371	0.396
GRPO [ 11 ]	0.410	0.380	0.389	0.405	0.382	0.443	0.409	0.387	0.397

TABLE III: Performance comparison: Lay Summarization. In this task, we follow [ 24 ] .

Methods	MediQ		ReDisQA		DiagnosisArena
Metrics	ACC	ROUGE-L	ACC	ROUGE-L	ACC	ROUGE-L
Qwen2.5-7B [ 7 ]	0.593	0.518	0.595	0.561	0.293	0.389
Deepseek-R1-7B [ 6 ]	0.358	0.450	0.389	0.257	0.250	0.163
Lingshu-7B [ 8 ]	0.589	0.467	0.566	0.529	0.423	0.164
SFT [ 9 ]	0.533	0.567	0.588	0.592	0.444	0.516
GRPO [ 11 ]	0.585	0.577	0.544	0.597	0.444	0.521

TABLE IV: Performance comparison: Clinical Diagnosis.

Algorithms	Metric	-NC	-NSW	-NLW	-NR	-NS	-NST	DSC
MedQA	ACC	0.190	0.210	0.280	0.260	0.240	0.210	0.290
eLife	ROUGE-1	0.419	0.420	0.439	0.432	0.427	0.427	0.448
	ROUGE-L	0.392	0.408	0.412	0.403	0.401	0.393	0.430
	SARI	0.424	0.451	0.449	0.451	0.450	0.433	0.453
DiagnosisArena	ACC	0.391	0.424	0.391	0.444	0.424	0.413	0.456
DiagnosisArena	ROUGE-L	0.465	0.479	0.518	0.523	0.508	0.491	0.532

TABLE V: Ablation Study. -NC ablates the retrieved context

\mathcal{C}

; -NSW/-NLW ablate the short/long entropy windows detection using fixed threshold, respectively; -NR ablates the

\mathcal{L}_{\text{rcf}}

regularization term; -NS ablates the entire Semantic Calibration Stream; and -NST ablates the Structure Calibration Stream.

Panel of Figure 9 — Fig. 4: Comparison under diverse retrievers. We employ the popular BMRETRIEVER [ 64 ] , E5 [ 62 ] , and MedCPT [ 42 ] . We compare retrieval-dependent baselines, including ICL, i-MedRAG, and TTT.

Panel of Figure 10 — Fig. 5: Comparison under diverse LLMs. We employ Qwen2.5-1.5B [ 7 ] , Lingshu-7B [ 8 ] , and Qwen3-14B [ 60 ] . We select ICL, SFT, i-MedRAG, and TTT as representative baselines. We additionally incorporate TAGS for MedQA and AgentSimp for eLife due to their competitive performance.

Panel of Figure 11 — Fig. 6: Comparison under diverse uncertainty estimation. Following [ 43 ] , we employ Entropy, Perplexity, and Energy. TTL-based baselines are included.

Panel of Figure 12 — Fig. 7: Online vs. Offline test-time optimization. Online methods tailor the model to each test query individually prior to inference, whereas offline methods conduct a one-time optimization on the entire test set before evaluating queries sequentially. TTL-based baselines are included.

Panel of Figure 13 — Fig. 8: OOD examination. (a) cross-dataset scenario. (b) cross-task scenario. For both scenarios, we use one domain as the pre-training / Index domain and then directly assess performance on the test set of the other domain.

Methods	MATH Reasoning		Knowledge QA-ReClor
Methods	AMC-ACC $\uparrow$	LogiQA-ACC $\uparrow$	ACC $\uparrow$	ROUGE-L $\uparrow$	SARI $\uparrow$
Qwen25-7B [ 7 ]	0.450	0.443	0.234	0.269	0.482
Linshu-7B [ 9 ]	0.040	0.388	0.205	0.259	0.479
ICL [ 17 ]	0.520	0.451	0.248	0.270	0.485
SFT [ 9 ]	0.500	0.488	0.246	0.227	0.466
Ensemble [ 61 ]	0.440	0.484	0.214	0.156	0.405

TABLE VI: Other Scenarios.

Panel of Figure 15 — Fig. 9: Time complexity. To demonstrate practicality and fairness, for Fig. 9(a) and 9(b) , we test the inference time for the eLife dataset on a machine equipped with one A800 GPU.

Panel of Figure 16 — Fig. 10: Case studies. Fig. 10(a) identifies critical tokens as those belonging to $\mathcal{U}$ (red) and certain tokens as those with $\mathbf{X}\setminus\mathcal{U}$ (blue). Fig. 10(b) illustrates the transition in predictive entropy during generation with and without entropy control. Fig. 10(c) displays the answering accuracy on auxiliary QA pairs (generated by DeepseekV3) derived from the context, comparing performance before and after the tuning process on eLife.

Fig. 11: Illustrative examples. The upper panel displays our response, while the lower panel shows the response generated by TTT. For visual clarity, we use placeholders for the questions and few-shot examples.

Panel of Figure 18 — Fig. 12: Hyper-parameter tests. Here, we take eLife as an example.

Clinical Reasoning

Clinical reasoning studies how AI systems interpret patient evidence, form and revise diagnoses, and support safe decisions across care settings.

See all 180 papers

TR-ICRL: Test-Time Rethinking for In-Context Reinforcement Learning
Apr 1, 2026Wenxuan Jiang, Yuxin Zuo, Zijian Zhang +8In-Context RL RL for Language Model Reasoning
The Hong Kong Polytechnic University · Meituan-M17 · Institute of Computing Technology, Chinese Academy of Sciences +3
Hit-RAG: Learning to Reason with Long Contexts via Preference Alignment
Mar 7, 2026Junming Liu, Yuqi Li, Shiping Wen +2Retrieval-Augmented Generation Multimodal RAG
Tongji University, Shanghai, China · The City University of New York, New York, USA · University of Technology Sydney, Sydney, Australia +2
Self-Improving In-Context Learning
May 22, 2026Baturay Saglam, Dionysis KalogeriasLanguage Model Calibration In-Context Learning
Department of Electrical and Computer Engineering Yale University

Apr 1, 2026J/K move · Enter open · S save

TR-ICRL: Test-Time Rethinking for In-Context Reinforcement Learning

Wenxuan Jiang, Yuxin Zuo, Zijian Zhang, Xuecheng Wu+7

The Hong Kong Polytechnic University · Meituan-M17 · Institute of Computing Technology, Chinese Academy of Sciences+3

In-Context RL RL for Language Model Reasoning Test-Time RL

Open paper AskPDF arXiv

In-Context Reinforcement Learning (ICRL) enables Large Language Models (LLMs) to learn online from external rewards directly within the context window. However, a central challenge in ICRL is reward estimation, as models typically lack access to ground-truths during inference. To address this limitation, we propose Test-Time Rethinking for In-Context Reinforcement Learning (TR-ICRL), a novel ICRL framework designed for both reasoning and knowledge-intensive tasks. TR-ICRL operates by first retrieving the most relevant instances from an unlabeled evaluation set for a given query. During each ICRL iteration, LLM generates a set of candidate answers for every retrieved instance. Next, a pseudo-label is derived from this set through majority voting. This label then serves as a proxy to give reward messages and generate formative feedbacks, guiding LLM through iterative refinement. In the end, this synthesized contextual information is integrated with the original query to form a comprehensive prompt, with the answer determining through a final round of majority voting. TR-ICRL is evaluated on mainstream reasoning and knowledge-intensive tasks, where it demonstrates significant performance gains. Remarkably, TR-ICRL improves Qwen2.5-7B by 21.23% on average on MedQA and even 137.59% on AIME2024. Extensive ablation studies and analyses further validate the effectiveness and robustness of our approach. Our code is available at https://github.com/pangpang-xuan/TR_ICRL.

Abstract

Clinical Reasoning

Explore similar work

TR-ICRL: Test-Time Rethinking for In-Context Reinforcement Learning

Hit-RAG: Learning to Reason with Long Contexts via Preference Alignment

Self-Improving In-Context Learning