Kevin Lin

@KevinQHLin

multimodal x agent x next postdoc @UniofOxford visiting @Stanford phd @NUSingapore | ex @Meta @Microsoft

Oxford, England

qhlin.me

Replies Reposts Media Articles

Pinned
Kevin Lin
@KevinQHLin
Jun 16
can AI write engaging news that people can trust? introducing ✨Data2Story: a data journalist agent. give it raw data, it generate a verifiable, multimodal article. 🔍verifiable: every claim is evidence-grounded, traces back to data, code, or a cited source. 🔮multimodal: the
00:00
12
Kevin Lin
@KevinQHLin
Sep 19
See H3 HyperFlow by Video Rebirth! Data-free flow self-distillation towards RSI on video-generation. github: github.com/Video-Rebirth/…
VIDEO REBIRTH
@video_rebirth
Sep 18
We're open-sourcing H3 HyperFlow. A data-free flow self-distillation technology built on @MiniMax_AI H3. No external training data. Significantly reduces inference cost while preserving frontier model quality. Full demo: videorebirth.com/lp/hyperflow
00:00
2
Kevin Lin
@KevinQHLin
Sep 10
See how computer-use agents solve challenging tasks through continuous trial and learning. GPT-6 Astra and Fable 5.1 already significantly outperform best human tester..
NeoCognition
@NeoCognition
Sep 10
Introducing ApprenticeBench: computer use + continual learning on a real job. We show Fable 5.1 and GPT-6 Astra can now continually learn on a job and surpass human professionals. A decisive step change in AI's job readiness. No FDEs. Agents deploy themselves into the job. 🧵
1
Kevin Lin
@KevinQHLin
Sep 10
the backpack boy coming again
Action Intelligence
@Action_Int_AI
Sep 10
Introducing #Continuo. Intelligence that stays with you. At #ECCV: understand the poster in front of you, remember what matters, and find what to explore next. Chapter 01 / FOR HUMAN From Action Intelligence, building robotic foundation models. @eccvconf
00:00
Kevin Lin
@KevinQHLin
Sep 10
🌟Check out our newest work Show-Harness 🕹️Towards Robot-Use for VLM, no calibration or extra VLA training needed. your VLM itself can play robot. 🦾 arxiv.org/abs/2609.10522 showlab.github.io/Show-Harness/
Zechen Bai
@ZechenBai
Sep 10
[1/8] 🦾 Could a strong robot policy already be hiding inside your VLM, not just GPT Astra, but also Gemini, Claude, even light-weight Qwen and InternVL? 🕹️ We are thrilled to introduce Show-Harness, an Embodied Harness that lets VLMs play real robots—like Claude Code for
00:00
2