影响力指数

76.59/100

前 1.7%

全站排名 #1,109

发表论文16 篇

平均评分5.7

年均产出5.3 篇/年

Ivan Vulić

Research Scientist@Google DeepMind·瑞士·OpenReview

研究方向

cross-lingual NLP · modularity · efficient ML · model merging · deep learning · semantic representations · multi-modal modeling · few-shot learning · transfer learning

6.0

Visual Planning: Let's Think Only with Images

ICLR 2026Oral

通讯

6.0

RAVENEA: A Benchmark for Multimodal Retrieval-Augmented Visual Culture Understanding

ICLR 2026Poster

5.3

Beyond the Final Layer: Intermediate Representations for Better Multilingual Calibration in Large Language Models

ICLR 2026Rejected

5.0

Multi-Agent Design: Optimizing Agents with Better Prompts and Topologies

ICLR 2026Poster

5.0

11Plus-Bench: Demystifying Multimodal LLM Spatial Reasoning with Cognitive-Inspired Analysis

ICLR 2026Rejected

2.7

Fleurs-SLU: A Massively Multilingual Benchmark for Spoken Language Understanding

COLM 2025Poster

二作

6.1

Multi-Agent Design: Optimizing Agents with Better Prompts and Topologies

ICML 2025Rejected

6.1

Imagine While Reasoning in Space: Multimodal Visualization-of-Thought

ICML 2025Poster

5.6

Measuring, Evaluating and Improving Logical Consistency in Large Language Models

ICLR 2025Rejected

4.9

Aligning with Logic: Measuring, Evaluating and Improving Logical Preference Consistency in Large Language Models

ICML 2025Spotlight

4.3

Language Fusion for Parameter-Efficient Cross-lingual Transfer

合作者 (20)

Ivan Vulić

Visual Planning: Let's Think Only with Images

RAVENEA: A Benchmark for Multimodal Retrieval-Augmented Visual Culture Understanding

Beyond the Final Layer: Intermediate Representations for Better Multilingual Calibration in Large Language Models

Multi-Agent Design: Optimizing Agents with Better Prompts and Topologies

11Plus-Bench: Demystifying Multimodal LLM Spatial Reasoning with Cognitive-Inspired Analysis

When Should AI Ask: Decision-theoretic Adaptive Communication for LLM Agents

Training Plug-and-Play Knowledge Modules with Deep Context Distillation

Universal Cross-Tokenizer Distillation via Approximate Likelihood Matching

Fleurs-SLU: A Massively Multilingual Benchmark for Spoken Language Understanding

Multi-Agent Design: Optimizing Agents with Better Prompts and Topologies

Imagine While Reasoning in Space: Multimodal Visualization-of-Thought

Measuring, Evaluating and Improving Logical Consistency in Large Language Models

Aligning with Logic: Measuring, Evaluating and Improving Logical Preference Consistency in Large Language Models

Language Fusion for Parameter-Efficient Cross-lingual Transfer