Wei-Hung Weng Profile
Wei-Hung Weng

@ckbjimmy

Followers
116
Following
39
Media
5
Statuses
16

Research Scientist @GoogleDeepMind, AI for science, MD/PhD, ex @MIT_CSAIL @HarvardDBMI

Joined April 2024
Don't wanna be here? Send us removal request.
@ckbjimmy
Wei-Hung Weng
22 days
AMIE can serve as a complimentary tool for ophthalmology. It not only matches specialist performance, but also augments ophthalmologist clinical reasoning, leading to improved accuracy, consensus and planning. [Preprint] https://t.co/T7x7W2cNx1
1
8
11
@ckbjimmy
Wei-Hung Weng
4 months
AMIE now can utilize an asynchronous oversight paradigm, managed by a team of specialized AI agents, guardrailed-AMIE (g-AMIE), to ensuree responsible and scalable human clinician oversight for medical AI. [Preprint] https://t.co/0Ea2PutowP
1
0
4
@ckbjimmy
Wei-Hung Weng
7 months
Powered by the natively multimodal Gemini and domain-specific inference-time algorithm (without further post-training!), AMIE can now see and interpret visual medical data within a diagnostic conversation [Preprint] https://t.co/yZfpyYRPBI [Blog] https://t.co/Mp0jO5ScYK
1
0
4
@ckbjimmy
Wei-Hung Weng
9 months
AMIE now gains enhanced management reasoning for disease management over time and longitudinal patient care using the Gemini-based dual-agentic system, with the integration of clinical knowledge! [Preprint] https://t.co/12kgweh1aT [Blog]
lnkd.in
This link will take you to a page that’s not on LinkedIn
@apalepu13
Anil Palepu
9 months
[1/N] I’m excited to share new work on Project AMIE (Articulate Medical Intelligence Explorer), our conversational medical AI! We put our system through a multi-visit OSCE to assess its management reasoning capabilities. Blog: https://t.co/RVkS5bd8NZ Paper:
1
0
9
@demishassabis
Demis Hassabis
9 months
Hypothesis generation and testing is a critical capability for AGI imo. Super excited about our AI co-scientist and other AI for Science work which are important steps towards that. We're on the cusp of an incredible new golden age of AI accelerated scientific discovery.
258
308
2K
@ckbjimmy
Wei-Hung Weng
9 months
Great work led by @jrpenades @CostaT_Lab, with our team @GoogleAI @GoogleDeepMind @googlecloud @alan_karthi @vivnat @Mysiak Annalisa Pawlosky Alexander Daryin @taotu831 @apalepu13 Artiom Myaskovsky
0
0
1
@ckbjimmy
Wei-Hung Weng
9 months
Pioneer researchers @jrpenades @CostaT_Lab (@imperialcollege) described how our AI co-scientist correctly recapitulated their unpublished, experimentally confirmed mechanism of bacterial gene transfer! Details https://t.co/tTdIvQa6YU
Tweet card summary image
biorxiv.org
Some mobile genetic elements spread among unrelated bacterial species through unknown mechanisms. Recently, we discovered that identical capsid-forming phage-inducible chromosomal islands (cf-PICIs),...
@alan_karthi
Alan Karthikesalingam
9 months
The reception to our @GoogleDeepMind @GoogleAI @googlecloud AI co-scientist has been amazing! We are working hard to expand access to the tool for more scientists. 🧫🧬A co-timed pre-print led by pioneer collaborators @imperialcollege, @jrpenades & @CostaT_Lab, describes how the
1
0
2
@ckbjimmy
Wei-Hung Weng
9 months
0
1
4
@ckbjimmy
Wei-Hung Weng
9 months
Delighted to introduce our collaborative, multi-agentic AI co-scientist to accelerate scientific discovery with scientists. Check out blog post https://t.co/1x9XBW8PLW and preprint https://t.co/eZctOk7I1d for signing up trusted tester program and more details!
@vivnat
Vivek Natarajan
9 months
Accelerating scientific discoveries and helping cure diseases might be the most profound purpose of AI. Thrilled to introduce our @GoogleAI @GoogleDeepMind @googlecloud AI co-scientist system, which I believe is an important milestone towards this. Blog -
1
0
9
@ckbjimmy
Wei-Hung Weng
1 year
Pleased to announce the release of the open-source, fully re-labeled MedQA benchmark, to facilitate more accurate model performance estimation. GitHub repository: https://t.co/ZgGHQnYGBS Check David's post for more details!
Tweet card summary image
github.com
For Med-Gemini, we relabeled the MedQA benchmark; this repo includes the annotations and analysis code. - Google-Health/med-gemini-medqa-relabelling
@davidstutz92
David Stutz
1 year
Part of our Med-Gemini work was a full relabelling of MedQA, revealing that at least 7.4% of examples are unfit for evaluation. Today, we open sourced these annotations alongside our evaluation script as a new standard evaluation on MedQA. A thread 🧵: https://t.co/7TE40sWUBU
0
1
7
@ckbjimmy
Wei-Hung Weng
2 years
Excited to share the work using conformal abstention to abstain LLM hallucinations!
@davidstutz92
David Stutz
2 years
Happy to share our latest work using conformal abstention to improve factuality of LLMs by avoiding hallucinations: https://t.co/i16SZYxLJp A short thread 🧵:
0
0
4
@ckbjimmy
Wei-Hung Weng
2 years
Excited to share Med-Gemini! https://t.co/E23dNAOgFR Except for remarkable SoTA performance on MedQA and various medical language/image/video tasks, one super exciting potential to highlight is to unlock novel applications via the long-context capability and multimodal dialogue!
@alan_karthi
Alan Karthikesalingam
2 years
Delighted to share ✨Med-Gemini✨ - our new family of multimodal models for medicine unlocking new possibilities for health - https://t.co/oemI52WBou More accurate multimodal conversations about medical images🩻, surgical videos📽️, genomics🧬, ultra-long health records📚, ECGs🫀
0
1
18