Cover Image

Finding the Translation Switch: Discovering and Exploiting the Task-Initiation Features in LLMs

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Wu, Xinwei, Liu, Heng, Zhao, Xiaohu, Ren, Yuqi, Xu, Linlong, Wang, Longyue, Xiong, Deyi, Luo, Weihua, Zhang, Kaifu
Format:	Preprint
Published:	2026
Subjects:	Computation and Language Artificial Intelligence
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Challenging Multilingual LLMs: A New Taxonomy and Benchmark for Unraveling Hallucination in Translation
by: Wu, Xinwei, et al.
Published: (2025)

Incentivizing Parametric Knowledge via Reinforcement Learning with Verifiable Rewards for Cross-Cultural Entity Translation
by: Zhou, Jiang, et al.
Published: (2026)

Beyond Single-Reward: Multi-Pair, Multi-Perspective Preference Optimization for Machine Translation
by: Wang, Hao, et al.
Published: (2025)

Rethinking Multilingual Vision-Language Translation: Dataset, Evaluation, and Adaptation
by: Wang, Xintong, et al.
Published: (2025)

(Perhaps) Beyond Human Translation: Harnessing Multi-Agent Collaboration for Translating Ultra-Long Literary Texts
by: Wu, Minghao, et al.
Published: (2024)

From Insight to Action: A Novel Framework for Interpretability-Guided Data Selection in Large Language Models
by: Shi, Ling, et al.
Published: (2026)

New Trends for Modern Machine Translation with Large Reasoning Models
by: Liu, Sinuo, et al.
Published: (2025)

TransBench: Benchmarking Machine Translation for Industrial-Scale Applications
by: Li, Haijun, et al.
Published: (2025)

DeepWideSearch: Benchmarking Depth and Width in Agentic Information Seeking
by: Lan, Tian, et al.
Published: (2025)

Towards Understanding Multi-Task Learning (Generalization) of LLMs via Detecting and Exploring Task-Specific Neurons
by: Leng, Yongqi, et al.
Published: (2024)

LHMKE: A Large-scale Holistic Multi-subject Knowledge Evaluation Benchmark for Chinese Large Language Models
by: Liu, Chuang, et al.
Published: (2024)

Marco-LLM: Bridging Languages via Massive Multilingual Training for Cross-Lingual Enhancement
by: Ming, Lingfeng, et al.
Published: (2024)

Exploring Multilingual Concepts of Human Value in Large Language Models: Is Value Alignment Consistent, Transferable and Controllable across Languages?
by: Xu, Shaoyang, et al.
Published: (2024)

ConTrans: Weak-to-Strong Alignment Engineering via Concept Transplantation
by: Dong, Weilong, et al.
Published: (2024)

Findings of the WMT 2024 Shared Task on Discourse-Level Literary Translation
by: Wang, Longyue, et al.
Published: (2024)

Do Large Language Models Mirror Cognitive Language Processing?
by: Ren, Yuqi, et al.
Published: (2024)

Towards Lightweight, Adaptive and Attribute-Aware Multi-Aspect Controllable Text Generation with Large Language Models
by: Zhu, Chenyu, et al.
Published: (2025)

LayAlign: Enhancing Multilingual Reasoning in Large Language Models via Layer-Wise Adaptive Fusion and Alignment Strategy
by: Ruan, Zhiwen, et al.
Published: (2025)

Marco-o1: Towards Open Reasoning Models for Open-Ended Solutions
by: Zhao, Yu, et al.
Published: (2024)

LANDeRMT: Detecting and Routing Language-Aware Neurons for Selectively Finetuning LLMs to Machine Translation
by: Zhu, Shaolin, et al.
Published: (2024)

The Bitter Lesson Learned from 2,000+ Multilingual Benchmarks
by: Wu, Minghao, et al.
Published: (2025)

VIDA: A dataset for Visually Dependent Ambiguity in Multimodal Machine Translation
by: Pan, Jingheng, et al.
Published: (2026)

An Empirical Study on the Robustness of Massively Multilingual Neural Machine Translation
by: Supryadi, et al.
Published: (2024)

ComfyUI-R1: Exploring Reasoning Models for Workflow Generation
by: Xu, Zhenran, et al.
Published: (2025)

Triplets Better Than Pairs: Towards Stable and Effective Self-Play Fine-Tuning for LLMs
by: Wang, Yibo, et al.
Published: (2026)

ComfyUI-Copilot: An Intelligent Assistant for Automated Workflow Development
by: Xu, Zhenran, et al.
Published: (2025)

HSCodeComp: A Realistic and Expert-level Benchmark for Deep Search Agents in Hierarchical Rule Application
by: Yang, Yiqian, et al.
Published: (2025)

A Unified Agentic Framework for Evaluating Conditional Image Generation
by: Wang, Jifang, et al.
Published: (2025)

Marco-Voice Technical Report
by: Tian, Fengping, et al.
Published: (2025)

The Box is in the Pen: Evaluating Commonsense Reasoning in Neural Machine Translation
by: He, Jie, et al.
Published: (2025)

Beyond Black-Box Interventions: Latent Probing for Faithful Retrieval-Augmented Generation
by: Gao, Linfeng, et al.
Published: (2025)

TagRouter: Learning Route to LLMs through Tags for Open-Domain Text Generation Tasks
by: Chen, Zhou, et al.
Published: (2025)

Efficiently Exploring Large Language Models for Document-Level Machine Translation with In-context Learning
by: Cui, Menglong, et al.
Published: (2024)

Can LLMs Track Their Output Length? A Dynamic Feedback Mechanism for Precise Length Regulation
by: Xiao, Meiman, et al.
Published: (2026)

AdaST: Dynamically Adapting Encoder States in the Decoder for End-to-End Speech-to-Text Translation
by: Huang, Wuwei, et al.
Published: (2025)

IRCAN: Mitigating Knowledge Conflicts in LLM Generation via Identifying and Reweighting Context-Aware Neurons
by: Shi, Dan, et al.
Published: (2024)

Marco-o1 v2: Towards Widening The Distillation Bottleneck for Reasoning Models
by: Yin, Huifeng, et al.
Published: (2025)

DCIS: Efficient Length Extrapolation of LLMs via Divide-and-Conquer Scaling Factor Search
by: Yang, Lei, et al.
Published: (2024)

TaP: A Taxonomy-Guided Framework for Automated and Scalable Preference Data Generation
by: Jin, Renren, et al.
Published: (2025)

Building Decision Making Models Through Language Model Regime
by: Zhang, Yu, et al.
Published: (2024)