mlsanityData to 14 Oct 2026

TraceICL: Thread-Resolved Demonstration Curation with Consistency Calibration for Robust Multimodal In-Context Learning

arXiv 2610.12516 ↗First posted 7 Oct 2026cs.CVcited by 0

On the map: Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning also fits: Image editing and quality assessment

Abstract (arXiv)

Interleaved multimodal in-context learning (ICL) lets frozen large vision-language models (LVLMs) acquire new tasks from a handful of image-text-answer demonstrations, yet it remains fragile: demonstrations bind to the wrong images, redundant exemplars crowd out informative ones, and noisy answers are copied verbatim into predictions. We present AVA, a training-free framework that organizes the demonstration pipeline of multimodal ICL into three composable stages. A Thread-Resolved Decomposer (TRD) segments the interleaved context into per-sample threads and re-stitches them with explicit boundaries, restoring image-text binding. An Influence-scored Demonstration Selector (IDS) estimates each candidate's marginal influence with one extra forward pass, removes redundancy, and orders the survivors by descending value. A Weak-to-Strong Consistency Calibration (WSC) module scores each demonstration by a weak sibling model's reconstruction confidence and softly reweights its contribution during decoding. On five vision-language benchmarks with interleaved 4-shot prompts, AVA improves the strongest same-setting baseline by +1.7 points on average, with gains of up to +2.4 points on VizWiz, shrinks the require-visual performance gap on ScienceQA-IMG, and retains most of its gain under 30% noisy demonstrations, at a 26% latency overhead that an offline cache amortizes. No gradients are ever computed; the LVLM stays frozen.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 35 of its 39 references that are in the map. First posted on arXiv 7 Oct 2026.

Builds on

35 of 39 references.

Closest ancestors · 5

Same direction, newest first

  1. 2023MMICL: Empowering Vision-language Model with Multi-Modal In-Context Learning14 Sep
  2. 2023Shikra: Unleashing Multimodal LLM's Referential Dialogue Magic27 Jun
  3. 2023MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models23 Jun
  4. 2023InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning11 May
  5. 2023Otter: A Multi-Modal Model with In-Context Instruction Tuning5 May

Other references · 30

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2023BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Modelscited by 6,137
  3. 2022Flamingo: a Visual Language Model for Few-Shot Learningcited by 4,386
  4. 2016Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answeringcited by 2,516
  5. 2022Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answeringcited by 1,919
  6. 2023Evaluating Object Hallucination in Large Vision-Language Modelscited by 1,729
  7. 2015Flickr30k Entities: Collecting Region-to-Phrase Correspondences for Richer Image-to-Sentence Modelscited by 1,362
  8. 2023OpenFlamingo: An Open-Source Framework for Training Large Autoregressive Vision-Language Modelscited by 565
  9. 2020The Hateful Memes Challenge: Detecting Hate Speech in Multimodal Memescited by 490
  10. 2023Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervisioncited by 467
  11. 2022Active Example Selection for In-Context Learningcited by 196
  12. 2022Self-Adaptive In-Context Learning: An Information Compression Perspective for In-Context Example Selection and Orderingcited by 161
18 more
  1. 2023Unified Demonstration Retriever for In-Context Learningcited by 142
  2. 2023Few-shot Fine-tuning vs. In-context Learning: A Fair Comparison and Evaluationcited by 136
  3. 2024Visual In-Context Learning for Large Vision-Language Modelscited by 128
  4. 2021WebSRC: A Dataset for Web-Based Structural Reading Comprehensioncited by 100
  5. 2022Language Models are General-Purpose Interfacescited by 97
  6. 2023Thread of Thought Unraveling Chaotic Contextscited by 72
  7. 2024ImgTrojan: Jailbreaking Vision-Language Models with ONE Imagecited by 57
  8. 2024GATEAU: Selecting Influential Samples for Long Context Alignmentcited by 20
  9. 2025A Goal Without a Plan Is Just a Wish: Efficient and Effective Global Planner Training for Long-Horizon Agent Taskscited by 19
  10. 2025DriVerse: Navigation World Model for Driving Simulation via Multimodal Trajectory Prompting and Motion Alignmentcited by 16
  11. 2025Teaching Large Language Models to Maintain Contextual Faithfulness via Synthetic Tasks and Reinforcement Learningcited by 11
  12. 2025U-ViLAR: Uncertainty-Aware Visual Localization for Autonomous Driving via Differentiable Association and Registrationcited by 5
  13. 2026Do Deployment Constraints Make LLMs Hallucinate Citations? An Empirical Study across Four Models and Five Prompting Regimescited by 3
  14. 2026Enhancing Multimodal In-Context Learning via Inductive-Deductive Reasoningcited by 2
  15. 2026When LLM Reward Design Fails: Diagnostic-Driven Refinement for Sparse Structured RLcited by 2
  16. 2026Confusion-Aware In-Context-Learning for Vision-Language Models in Robotic Manipulationcited by 1
  17. 2026Parallel In-context Learning for Large Vision Language Modelscited by 1
  18. 2026Geometry-Aware Uncertainty Coresets for Robust Visual In-Context Learning in Histopathologycited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Advancing Multimodal In-Context Learning in Large Vision-Language Models with Task-aware Demonstrations
  2. 2025What do vision-language models see in the context? Investigating multimodal in-context learning
  3. 2026Hyper-ICL: Attention Calibration with Hyperbolic Anchor Distillation for Multimodal In-Context Learning
  4. 2024From Introspection to Best Practices: Principled Analysis of Demonstrations in Multimodal In-Context Learning
  5. 2024Towards Multimodal In-Context Learning for Vision & Language Models
  6. 2026In-Context Collapse in Vision-Language Models and How to Mitigate it?
  7. 2025M2IV: Towards Efficient and Fine-grained Multimodal In-Context Learning via Representation Engineering
  8. 2025Mimicking or Reasoning: Rethinking Multi-Modal In-Context Learning in Vision-Language Models