Shulin Tian

@shulin_tian

Ph.D. Student @MMLabNTU + ACIS Scholar @ASTARsg | Agents, Vision & Learning

Singapore

shulin16.github.io

Joined October 2021

326 Following

293 Followers

Replies Reposts Media

Pinned
Shulin Tian
@shulin_tian
Oct 3
Our ancestors picked up a rock. They saw a hammer, a blade, and a way to survive. Humans didn’t conquer the world with bare hands -- we did it with tools. 🛠️ Behind every tool we reach for is a tiny act of intelligence: "What do I need? What should I use? What do I do next?"
6
Shulin Tian
@shulin_tian
Oct 26
I will be presenting V-Rubrics and EgoTools at #EMNLP2026 in Budapest 🇭🇺 Monday, Oct 26 · Hall H, HUNGEXPO • V-Rubrics: 09:00–10:30 • EgoTools: 16:45–18:15 Please feel free to stop by for a chat ;)
Shulin Tian
@shulin_tian
Oct 7
DiVeR improves test-time scaling for vision-language-action (VLA) models by training verifiers to focus on 𝘀𝘁𝗮𝘁𝗲𝘀 𝘄𝗵𝗲𝗿𝗲 𝗮𝗰𝘁𝗶𝗼𝗻 𝗰𝗵𝗼𝗶𝗰𝗲𝘀 𝗺𝗮𝘁𝘁𝗲𝗿 𝗺𝗼𝘀𝘁. TL;DR: DiVeR estimates decision criticality from the variance of sampled action representations,
This post is unavailable.
Shulin Tian
@shulin_tian
Oct 6
I will be giving a talk on our recent paper, V-rubrics, on October 8 at 8PM (GMT+8). Please feel free to PM me for the link ;)
青稞社区
@qingke_ai
Oct 4
北京时间10月8日（周四）晚8点，#青稞Talk第157期，#青稞社区邀请到南洋理工大学计算机科学博士生@shulin_tian 田淑琳直播分享V-Rubrics 包括 V-Rubrics 50K
1
Shulin Tian
@shulin_tian
Aug 27
Vision-language models can produce fluent, plausible answers while still getting a visual fact or intermediate inference wrong. We view this as a credit-assignment problem: a single outcome reward may tell us whether the final answer is acceptable, but it cannot identify which

Shulin Tian

Shulin Tian’s posts