mlsanityData to 7 Oct 2026

Calibrate the Decisions That Change the Future: On-Policy Post-Training Quantization for Multimodal Large Language Models

arXiv 2609.36828 ↗First posted 29 Sep 2026cs.AIcited by 0

On the map: Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning also fits: Embodied and 3D vision-language models

Abstract (arXiv)

Post-training quantization (PTQ) lowers deployment cost for multimodal large language models, but calibration typically reconstructs fixed sequences with local objectives. This overlooks autoregressive feedback: a quantization-induced token change redirects the prefix and changes future states. Yet on-policy coverage alone is insufficient because many decision mismatches barely affect future generation. We propose OnPTQ, an on-policy framework that calibrates on trajectories visited by the current quantized policy. On shared prefixes, OnPTQ identifies quantization-eroded boundaries, evaluates competing tokens through short counterfactual rollouts, and combines current discrepancy with branch consequence into a Decision--Consequence risk. The risk prioritizes critical states, while context anchoring and trajectory refresh preserve multimodal behavior and keep calibration aligned with the updated policy. We further derive a Decision--Consequence bound linking behavioral deviation to current policy discrepancy and action-conditioned future-value span. Across vision--language and omni-modal Qwen models under multiple low-bit settings, OnPTQ improves downstream performance and yields fewer correctness flips against the corresponding Dense/FP16 references, without changing the deployed inference graph.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 25 of its 33 references that are in the map. First posted on arXiv 29 Sep 2026.

Builds on

25 of 33 references.

Closest ancestors · 1

Same direction, newest first

  1. 2026MASQuant: Modality-Aware Smoothing Quantization for Multimodal Large Language Models5 Mar

Other references · 24

  1. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,044
  2. 2022Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answeringcited by 1,881
  3. 2022GPTQ: Accurate Post-Training Quantization for Generative Pre-trained Transformerscited by 1,838
  4. 2019Towards VQA Models That Can Readcited by 1,696
  5. 2022SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Modelscited by 1,338
  6. 2023AWQ: Activation-aware Weight Quantization for LLM Compression and Accelerationcited by 1,234
  7. 2018VizWiz Grand Challenge: Answering Visual Questions from Blind Peoplecited by 852
  8. 2015Scheduled Sampling for Sequence Prediction with Recurrent Neural Networkscited by 779
  9. 2024QuaRot: Outlier-Free 4-Bit Inference in Rotated LLMscited by 598
  10. 2015Sequence Level Training with Recurrent Neural Networkscited by 568
  11. 2022ZeroQuant: Efficient and Affordable Post-Training Quantization for Large-Scale Transformerscited by 517
  12. 2023OCRBench: On the Hidden Mystery of OCR in Large Multimodal Modelscited by 466
12 more
  1. 2024SpinQuant: LLM quantization with learned rotationscited by 446
  2. 2023OmniQuant: Omnidirectionally Calibrated Quantization for Large Language Modelscited by 403
  3. 2016Professor Forcing: A New Algorithm for Training Recurrent Networkscited by 214
  4. 2021WenetSpeech: A 10000+ Hours Multi-domain Mandarin Corpus for Speech Recognitioncited by 189
  5. 2024FlatQuant: Flatness Matters for LLM Quantizationcited by 116
  6. 2020Autoregressive Knowledge Distillation through Imitation Learningcited by 68
  7. 2024Q-VLM: Post-training Quantization for Large Vision-Language Modelscited by 47
  8. 2024MBQ: Modality-Balanced Quantization for Large Vision-Language Modelscited by 45
  9. 2025GuidedQuant: Large Language Model Quantization via Exploiting End Loss Guidancecited by 28
  10. 2024LRQ: Optimizing Post-Training Quantization for Large Language Models by Learning Low-Rank Weight-Scaling Matricescited by 11
  11. 2026Fine-Grained Post-Training Quantization for Large Vision Language Models with Quantization-Aware Integrated Gradientscited by 7
  12. 2026Quant Experts: Token-aware Adaptive Error Reconstruction with Mixture of Experts for Large Vision-Language Models Quantizationcited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Large Language Model Post-Training: A Unified View of Off-Policy and On-Policy Learning
  2. 2026Back on Track: Aligning Rewards and States for Reasoning in Diffusion Large Language Models
  3. 2026Multi-Branch Policy Optimization for Multimodal Large Language Models
  4. 2026Reinforcement Inference: Leveraging Uncertainty for Self-Correcting Language Model Reasoning
  5. 2026Predict Before You Deploy: Offline Prediction of Quantization-Induced Task Degradation for World Action Models
  6. 2026Beyond the Stability-Exploration Dilemma: Environmental Regularization for LLM Policy Optimization
  7. 2026Balancing Classification and Calibration Performance in Decision-Making LLMs via Calibration Aware Reinforcement Learning
  8. 2026Beyond the Current Observation: Evaluating Multimodal Large Language Models in Controllable Non-Markov Games