Cover Image

Pragma-VL: Towards a Pragmatic Arbitration of Safety and Helpfulness in MLLMs

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wen, Ming, Yang, Kun, Chen, Xin, Zhang, Jingyu, Han, Dingding, Cui, Shiwen, Xu, Yuedong
Format:	Preprint
Published:	2026
Subjects:	Machine Learning Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Differentially Private Federated Low Rank Adaptation Beyond Fixed-Matrix
by: Wen, Ming, et al.
Published: (2025)

UpSafe$^\circ$C: Upcycling for Controllable Safety in Large Language Models
by: Sun, Yuhao, et al.
Published: (2025)

VL Norm: Rethink Loss Aggregation in RLVR
by: He, Zhiyuan, et al.
Published: (2025)

The Forgotten Shield: Safety Grafting in Parameter-Space for Medical MLLMs
by: Zhao, Jiale, et al.
Published: (2025)

CSR-Bench: A Benchmark for Evaluating the Cross-modal Safety and Reliability of MLLMs
by: Liu, Yuxuan, et al.
Published: (2026)

From Specific-MLLMs to Omni-MLLMs: A Survey on MLLMs Aligned with Multi-modalities
by: Jiang, Shixin, et al.
Published: (2024)

OOD-MMSafe: Advancing MLLM Safety from Harmful Intent to Hidden Consequences
by: Wen, Ming, et al.
Published: (2026)

TimeOmni-VL: Unified Models for Time Series Understanding and Generation
by: Guan, Tong, et al.
Published: (2026)

Can Large Language Models Help Experimental Design for Causal Discovery?
by: Li, Junyi, et al.
Published: (2025)

Exploring Response Uncertainty in MLLMs: An Empirical Evaluation under Misleading Scenarios
by: Dang, Yunkai, et al.
Published: (2024)

Support Sufficiency as Consequence-Sensitive Compression in Belief Arbitration
by: Walsh, Mark
Published: (2026)

VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learning
by: Wang, Haozhe, et al.
Published: (2025)

Urban-R1: Reinforced MLLMs Mitigate Geospatial Biases for Urban General Intelligence
by: Wang, Qiongyan, et al.
Published: (2025)

De Novo Molecular Generation from Mass Spectra via Many-Body Enhanced Diffusion
by: Sun, Xichen, et al.
Published: (2026)

From Pragmas to Partners: A Symbiotic Evolution of Agentic High-Level Synthesis
by: Zhang, Niansong, et al.
Published: (2026)

NVIDIA Nemotron Nano V2 VL
by: NVIDIA, et al.
Published: (2025)

Bi-Factorial Preference Optimization: Balancing Safety-Helpfulness in Language Models
by: Zhang, Wenxuan, et al.
Published: (2024)

DeepSeek-VL: Towards Real-World Vision-Language Understanding
by: Lu, Haoyu, et al.
Published: (2024)

Pelican-VL 1.0: A Foundation Brain Model for Embodied Intelligence
by: Zhang, Yi, et al.
Published: (2025)

Towards Effective Experiential Learning: Dual Guidance for Utilization and Internalization
by: Bai, Fei, et al.
Published: (2026)

Discriminative classification with generative features: bridging Naive Bayes and logistic regression
by: Terner, Zachary, et al.
Published: (2025)

VL-KGE: Vision-Language Models Meet Knowledge Graph Embeddings
by: Efthymiou, Athanasios, et al.
Published: (2026)

A Safety Report on GPT-5.2, Gemini 3 Pro, Qwen3-VL, Grok 4.1 Fast, Nano Banana Pro, and Seedream 4.5
by: Ma, Xingjun, et al.
Published: (2026)

Towards Robust Offline Reinforcement Learning under Diverse Data Corruption
by: Yang, Rui, et al.
Published: (2023)

Learning Safety Constraints for Large Language Models
by: Chen, Xin, et al.
Published: (2025)

Towards Few-shot Self-explaining Graph Neural Networks
by: Peng, Jingyu, et al.
Published: (2024)

Mix Data or Merge Models? Balancing the Helpfulness, Honesty, and Harmlessness of Large Language Model via Model Merging
by: Yang, Jinluan, et al.
Published: (2025)

Early Rug Pull Warning for BSC Meme Tokens via Multi-Granularity Wash-Trading Pattern Profiling
by: Cao, Dingding, et al.
Published: (2026)

Dirichlet-Based Prediction Calibration for Learning with Noisy Labels
by: Zong, Chen-Chen, et al.
Published: (2024)

A Communication and Computation Efficient Fully First-order Method for Decentralized Bilevel Optimization
by: Wen, Min, et al.
Published: (2024)

Pragmatic Policy Development via Interpretable Behavior Cloning
by: Matsson, Anton, et al.
Published: (2025)

VL-RouterBench: A Benchmark for Vision-Language Model Routing
by: Huang, Zhehao, et al.
Published: (2025)

COSMO-RL: Towards Trustworthy LMRMs via Joint Safety and Stability
by: Ding, Yizhuo, et al.
Published: (2025)

Risk Awareness Injection: Calibrating Vision-Language Models for Safety without Compromising Utility
by: Wang, Mengxuan, et al.
Published: (2026)

Advancing Deep Learning through Probability Engineering: A Pragmatic Paradigm for Modern AI
by: Zhang, Jianyi
Published: (2025)

Learning Identifiable Structures Helps Avoid Bias in DNN-based Supervised Causal Learning
by: Zhang, Jiaru, et al.
Published: (2025)

Towards Monotonic Improvement in In-Context Reinforcement Learning
by: Zhang, Wenhao, et al.
Published: (2025)

SafeSeek: Universal Attribution of Safety Circuits in Language Models
by: Yu, Miao, et al.
Published: (2026)

Player-Team Heterogeneous Interaction Graph Transformer for Soccer Outcome Prediction
by: Wang, Lintao, et al.
Published: (2025)

OptProver: Bridging Olympiad and Optimization through Continual Training in Formal Theorem Proving
by: Li, Chenyi, et al.
Published: (2026)