Patrick Amadeus Irawan

MBZUAI

Papers on Lattice

Total citations

Topics

h-index

Research focus

Multimodal Models (3)Eval Frameworks & Benchmarks (2)Constitutional AI & AI Ethics (1)Reasoning & Chain-of-Thought (1)Inference & Quantization (1)

Frequent co-authors

Alham Fikri Aji (2)Samuel Cahyawijaya (1)Peerat Limkonchotiwat (1)Tack Hwa Wong (1)

Papers (3)

Apr 13, 2026

NUSApr 13, 2026·also Brown, De La Salle University, MBZUAI, National University +1

Anthropogenic Regional Adaptation in Multimodal Vision-Language Model

Fine-tuning VLMs for regional relevance doesn't have to sacrifice global performance: a simple data filtering and model merging technique boosts cultural relevance by 5-15% while barely impacting overall accuracy.

Samuel Cahyawijaya, Peerat Limkonchotiwat, Tack Hwa Wong +50

Constitutional AI & AI Ethics Eval Frameworks & Benchmarks Multimodal Models

Apr 11, 2026

Counting to Four is still a Chore for VLMs

VLMs struggle with basic counting not because they can't "see" the objects, but because they forget to look when generating the answer.

Duy Le Dinh Anh, Patrick Amadeus Irawan, T. Vo

Eval Frameworks & Benchmarks Multimodal Models Reasoning & Chain-of-Thought

Apr 1, 2026

LinguDistill: Recovering Linguistic Ability in Vision- Language Models via Selective Cross-Modal Distillation

VLMs can regain lost linguistic prowess without extra parameters or architectural changes, thanks to a clever KV-cache sharing trick for distillation.

Patrick Amadeus Irawan, Erland Hilman Fuadi, Shanu Kumar +2

Inference & Quantization Multimodal Models Natural Language Processing+1

Search

Patrick Amadeus Irawan

Research focus

Frequent co-authors

Papers (3)