01 / PERSONAL PROFILE
RESEARCH · INDUSTRY · SCIENCE

HELLO, I'M

Zhihao
L李志豪

RESEARCH SCIENTIST @ BOSCH CORPORATE RESEARCH

Building AI that understands the physical world — and improves how we engineer it.

My research explores AI for Science, Industrial AI, AI Agents, and Foundation Models, connecting computational intelligence with real-world engineering challenges.

Black-and-white editorial portrait of Zhihao LI with round glasses
© 2026 / PROFILE
01 / CURRENTResearch Scientist
Bosch Corporate Research
02 / EDUCATIONPh.D. · Data Science & Analytics
HKUST (Guangzhou)
03 / RESEARCH INTERESTSAI for Science · Industrial AI
Agents · Foundation Models

What I work on.

LEARN · SIMULATE · EVOLVE

I am interested in AI systems that represent physical processes, learn efficiently from scientific data, and support autonomous engineering workflows. These themes are closely connected rather than independent research silos.

R / 01↗

Self-Evolving Agentic Engineering

Self-evolving engineering agents that plan experiments, accumulate verified knowledge, and continuously improve models and decisions across tasks.

# AI AGENTS # SELF-EVOLUTION
R / 02↗

Scientific Foundation Models

Reusable representations and data-efficient learning across scientific tasks and physical systems.

# FOUNDATION MODELS  # AI FOR SCIENCE
R / 03↗

Physics-Aware Learning

Machine learning architectures that exploit physical structure for accurate, efficient simulation.

# PHYSICAL AI  # EFFICIENT LEARNING

Publications.

THE WORK MATTERS,
NOT THE COUNT.
FIG. / GPO
ICML 2026Interpretable AINeural OperatorsSeoul, South KoreaJul 6–11, 2026

From Basis to Basis: Gaussian Particle Representation for Interpretable PDE Operators

Zhihao LI, Yu Feng, Zhilu Lai, Wei Wang

TL;DR  A Gaussian-particle basis makes neural operators interpretable and resolution-agnostic, with efficient coupling in compact modal space.

PaperPDFCode
FIG. / REMD
CVPR 2026Fluid SRDiffusionDenver, USAJun 3–7, 2026

Physics-Consistent Diffusion for Efficient Fluid Super-Resolution via Multiscale Residual Correction

Zhihao LI, Shengwei Dong, Chuang Yi, Junxuan Gao, Zhilu Lai, Zhiqiang Liu, Wei Wang, Guangtao Zhang

TL;DR  ReMD integrates multiscale residual correction and physics consistency inside the diffusion process for efficient fluid super-resolution.

PaperPDFCode
FIG. / M2NO
KDD 2026Multi-ResolutionPDE SolversJeju, South KoreaAug 9–13, 2026

M2NO: An Efficient Multi-Resolution Operator Framework for Dynamic Multi-Scale PDE Solvers

Zhihao LI, Zhilu Lai, Xiaobo Zhang, Wei Wang

TL;DR  A multi-resolution neural operator combines hierarchical PDE representations and multiscale modeling for efficient dynamic simulations.

PaperPDFCode
FIG. / AMG
KDD 2025Arbitrary GeometriesMulti-GraphToronto, CanadaAug 3–7, 2025

Harnessing Scale and Physics: A Multi-Graph Neural Operator Framework for PDEs on Arbitrary Geometries

Zhihao LI, Haoze Song, Di Xiao, Zhilu Lai, Wei Wang

TL;DR  Multi-graph neural operators fuse geometry, physical relations, and scale for PDE learning on arbitrary domains.

PaperPDFCode

Latest news.

2026.09
One paper accepted at NeurIPS 2026

Reformulating Neural Operators in d+1 Dimensions for Embedding Evolution.

2026.09
Joined Bosch Corporate Research

Research Scientist · AI for Science and Industrial AI.

2026.05
One paper accepted at ICML 2026

From Basis to Basis: Gaussian Particle Representation for Interpretable PDE Operators.

2026.05
Successfully defended my Ph.D. dissertation

Towards Scalable and Generalizable Neural PDE Solvers: Learning Across Scales, Geometries, and Physics

2026.02
One paper accepted at CVPR 2026

Physics-Consistent Diffusion for Efficient Fluid Super-Resolution via Multiscale Residual Correction.

Background.

Bosch Corporate Research

Research Scientist

2026 — PRESENT
Huawei 2012 Laboratories

Research Intern

2025 — 2026
Hong Kong University of Science and Technology (Guangzhou)

Ph.D. · Data Science & Analytics

2022 — 2026
Supervisor: Wei WangCo-supervisor: Zhilu Lai
Harbin Institute of Technology, Shenzhen

B.Eng. · Computer Science & Technology

2018 — 2022
Supervisor: Qing Liao
06 / CONNECT & COLLABORATE

Let's build something
meaningful.

Interested in research collaborations or scientific AI for engineering? I would be glad to connect.

Get in touch ↗
(() => { const toast = document.getElementById('toast'); let timer; function notify(message) {toast.textContent=message;toast.classList.add('show');clearTimeout(timer);timer=setTimeout(()=>toast.classList.remove('show'),3000)} const menu = document.getElementById('mobile-menu'); const nav = document.getElementById('navigation'); menu.addEventListener('click',()=>{const open=nav.classList.toggle('open');menu.setAttribute('aria-expanded',open);menu.setAttribute('aria-label',open?'Close navigation menu':'Open navigation menu');menu.textContent=open?'×':'☰'}); nav.querySelectorAll('a').forEach(link=>link.addEventListener('click',()=>{nav.classList.remove('open');menu.setAttribute('aria-expanded','false');menu.textContent='☰'})); document.querySelectorAll('[data-unconfigured]').forEach(link=>link.addEventListener('click',event=>{event.preventDefault();notify('Preview only: add your real '+link.dataset.unconfigured+' before publishing.')})); const filterButtons=document.querySelectorAll('[data-filter]'); const papers=document.querySelectorAll('.paper'); filterButtons.forEach(btn=>btn.addEventListener('click',()=>{const choice=btn.dataset.filter;filterButtons.forEach(b=>b.setAttribute('aria-pressed',String(b===btn)));papers.forEach(p=>{p.hidden=choice==='selected'?p.dataset.selected!=='true':choice==='all'?false:p.dataset.category!==choice})})); papers.forEach(p=>{p.querySelectorAll('[data-action="abstract"],[data-action="bib"]').forEach(btn=>btn.addEventListener('click',()=>{const detail=p.querySelector('[data-detail="'+btn.dataset.action+'"]'); const open=detail.hidden;detail.hidden=!open;btn.setAttribute('aria-expanded',String(open));btn.textContent=(btn.dataset.action==='bib'?'BibTex':'Abstract')})); const copy=p.querySelector('[data-action="copy"]');copy.addEventListener('click',async()=>{const txt=p.querySelector('[data-detail="bib"] pre').innerText;try{await navigator.clipboard.writeText(txt);notify('BibTex copied to clipboard.')}catch{const range=document.createRange();range.selectNodeContents(p.querySelector('[data-detail="bib"] pre'));const sel=window.getSelection();sel.removeAllRanges();sel.addRange(range);notify('Select and copy the highlighted BibTex.')}})}); papers.forEach(p=>{ function toggleHighlight(){const was=p.dataset.highlighted==='true';papers.forEach(x=>x.dataset.highlighted='false');p.dataset.highlighted=String(!was)} p.addEventListener('click',event=>{if(event.target.closest('a,button'))return;toggleHighlight()}); p.addEventListener('keydown',event=>{if((event.key==='Enter'||event.key===' ')&&event.target===p){event.preventDefault();toggleHighlight()}}) }); if('IntersectionObserver' in window){const links=[...document.querySelectorAll('.nav a[href^="#"]')];const observer=new IntersectionObserver(entries=>{entries.forEach(entry=>{if(entry.isIntersecting){links.forEach(link=>link.classList.toggle('current',link.getAttribute('href')==='#'+entry.target.id))}})},{rootMargin:'-20% 0px -65% 0px'});['about','research','publications','news','experience'].forEach(id=>{const e=document.getElementById(id);if(e)observer.observe(e)})} })();