Copertina

Unlocking Multilingual Reasoning Capability of LLMs and LVLMs through Representation Engineering

Fuente: arXiv

Salvato in:

Dettagli Bibliografici
Autori principali:	Li, Qiming, Feng, Xiaocheng, Ma, Yixuan, Ye, Zekai, Chen, Ruihan, Feng, Xiachong, Qin, Bing
Natura:	Preprint
Pubblicazione:	2025
Soggetti:	Computer Vision and Pattern Recognition
Accesso online:	Acceder al recurso
Tags:	Aggiungi Tag Nessun Tag, puoi essere il primo ad aggiungerne!!

Documenti analoghi

Causal Tracing of Object Representations in Large Vision Language Models: Mechanistic Interpretability and Hallucination Mitigation
di: Li, Qiming, et al.
Pubblicazione: (2025)

CAI: Caption-Sensitive Attention Intervention for Mitigating Object Hallucination in Large Vision-Language Models
di: Li, Qiming, et al.
Pubblicazione: (2025)

CAST: Mitigating Object Hallucination in Large Vision-Language Models via Caption-Guided Visual Attention Steering
di: Li, Qiming, et al.
Pubblicazione: (2026)

CLAIM: Mitigating Multilingual Object Hallucination in Large Vision-Language Models with Cross-Lingual Attention Intervention
di: Ye, Zekai, et al.
Pubblicazione: (2025)

Unlocking Few-Shot Capabilities in LVLMs via Prompt Conditioning and Head Selection
di: de Senneville, Adhemar, et al.
Pubblicazione: (2026)

Cross-Lingual Text-Rich Visual Comprehension: An Information Theory Perspective
di: Yu, Xinmiao, et al.
Pubblicazione: (2024)

Self-Prophetic Decoding to Unlock Visual Search in LVLMs
di: He, Zhendong, et al.
Pubblicazione: (2026)

CrossWordBench: Evaluating the Reasoning Capabilities of LLMs and LVLMs with Controllable Puzzle Generation
di: Leng, Jixuan, et al.
Pubblicazione: (2025)

OMIBench: Benchmarking Olympiad-Level Multi-Image Reasoning in Large Vision-Language Model
di: Chen, Qiguang, et al.
Pubblicazione: (2026)

Knowing the Answer Isn't Enough: Fixing Reasoning Path Failures in LVLMs
di: Wang, Chaoyang, et al.
Pubblicazione: (2025)

Investigating and Mitigating the Multimodal Hallucination Snowballing in Large Vision-Language Models
di: Zhong, Weihong, et al.
Pubblicazione: (2024)

Elevating Visual Question Answering through Implicitly Learned Reasoning Pathways in LVLMs
di: Jing, Liu, et al.
Pubblicazione: (2025)

See It, Say It, Sorted: An Iterative Training-Free Framework for Visually-Grounded Multimodal Reasoning in LVLMs
di: Zhang, Yongchang, et al.
Pubblicazione: (2026)

Masking Matters: Unlocking the Spatial Reasoning Capabilities of LLMs for 3D Scene-Language Understanding
di: Jeon, Yerim, et al.
Pubblicazione: (2025)

Unlocking the Capabilities of Large Vision-Language Models for Generalizable and Explainable Deepfake Detection
di: Yu, Peipeng, et al.
Pubblicazione: (2025)

Mitigating Action-Relation Hallucinations in LVLMs via Relation-aware Visual Enhancement
di: Qin, Zhenxin, et al.
Pubblicazione: (2026)

Unleashing the Intrinsic Visual Representation Capability of Multimodal Large Language Models
di: Li, Hengzhuang, et al.
Pubblicazione: (2025)

EditThinker: Unlocking Iterative Reasoning for Any Image Editor
di: Li, Hongyu, et al.
Pubblicazione: (2025)

SAVAA: Mitigating Hallucinations in LVLMs via Step-wise Adaptive Visual Attention Amplification
di: Zhang, Jiacheng, et al.
Pubblicazione: (2026)

Interpretable Oracle Bone Script Decipherment through Radical and Pictographic Analysis with LVLMs
di: Peng, Kaixin, et al.
Pubblicazione: (2025)

Mitigating Object Hallucinations in LVLMs via Attention Imbalance Rectification
di: Sun, Han, et al.
Pubblicazione: (2026)

From Redundancy to Relevance: Information Flow in LVLMs Across Reasoning Tasks
di: Zhang, Xiaofeng, et al.
Pubblicazione: (2024)

VAT: Vision Action Transformer by Unlocking Full Representation of ViT
di: Li, Wenhao, et al.
Pubblicazione: (2025)

DianJin-OCR-R1: Enhancing OCR Capabilities via a Reasoning-and-Tool Interleaved Vision-Language Model
di: Chen, Qian, et al.
Pubblicazione: (2025)

Causally-Grounded Dual-Path Attention Intervention for Object Hallucination Mitigation in LVLMs
di: Yu, Liu, et al.
Pubblicazione: (2025)

ShredBench: Evaluating the Semantic Reasoning Capabilities of Multimodal LLMs in Document Reconstruction
di: Guo, Zichun, et al.
Pubblicazione: (2026)

IVC-Prune: Revealing the Implicit Visual Coordinates in LVLMs for Vision Token Pruning
di: Sun, Zhichao, et al.
Pubblicazione: (2026)

Evidence Packing for Cross-Domain Image Deepfake Detection with LVLMs
di: Liu, Yuxin, et al.
Pubblicazione: (2026)

The Geometry of Compromise: Unlocking Generative Capabilities via Controllable Modality Alignment
di: Liu, Hongyuan, et al.
Pubblicazione: (2026)

Long-CLIP: Unlocking the Long-Text Capability of CLIP
di: Zhang, Beichen, et al.
Pubblicazione: (2024)

MECD: Unlocking Multi-Event Causal Discovery in Video Reasoning
di: Chen, Tieyuan, et al.
Pubblicazione: (2024)

Vocabulary Hijacking in LVLMs: Unveiling Critical Attention Heads by Excluding Inert Tokens to Mitigate Hallucination
di: Chen, Yangneng, et al.
Pubblicazione: (2026)

HELPD: Mitigating Hallucination of LVLMs by Hierarchical Feedback Learning with Vision-enhanced Penalty Decoding
di: Yuan, Fan, et al.
Pubblicazione: (2024)

VQTalker: Towards Multilingual Talking Avatars through Facial Motion Tokenization
di: Liu, Tao, et al.
Pubblicazione: (2024)

LVLMs as inspectors: an agentic framework for category-level structural defect annotation
di: Jiang, Sheng, et al.
Pubblicazione: (2025)

Benchmarking Corruption Robustness of LVLMs: A Discriminative Benchmark and Robustness Alignment Metric
di: Sui, Xiangjie, et al.
Pubblicazione: (2025)

CoV: Chain-of-View Prompting for Spatial Reasoning
di: Zhao, Haoyu, et al.
Pubblicazione: (2026)

ContextGuard-LVLM: Enhancing News Veracity through Fine-grained Cross-modal Contextual Consistency Verification
di: Ma, Sihan, et al.
Pubblicazione: (2025)

TACO: Think-Answer Consistency for Optimized Long-Chain Reasoning and Efficient Data Learning via Reinforcement Learning in LVLMs
di: Kan, Zhehan, et al.
Pubblicazione: (2025)

See Different, Think Better: Visual Variations Mitigating Hallucinations in LVLMs
di: Dai, Ziyun, et al.
Pubblicazione: (2025)