praveersingh90 Profile Banner
Praveer Singh Profile
Praveer Singh

@praveersingh90

Followers
325
Following
2K
Media
26
Statuses
574

Assistant Professor @CUMedicalSchool | @CUEyeCenter @CUAnschutz | Honorary Researcher: @Moorfields Past: @harvardmed @MGHImaging @MGHMartinos

Colorado, USA
Joined August 2013
Don't wanna be here? Send us removal request.
@docmilanfar
Peyman Milanfar
6 months
You should be so lucky to have people throughout your research career that you can openly bounce ideas to and from - especially if they complement your strengths in your areas of weakness - it is a rare and precious gift.
15
99
738
@WHO
World Health Organization (WHO)
9 months
Women shouldn’t require superpowers to juggle breastfeeding and work. All moms everywhere, no matter their work or contract type, should have: 🧸 At least 18 weeks, preferably 6 months or more, of paid maternity leave 🧸 Paid time off for breastfeeding or expressing milk upon
30
107
274
@robertarail
Roberta Raileanu
9 months
Super excited to share 🧠MLGym 🦾 – the first Gym environment for AI Research Agents 🤖🔬 We introduce MLGym and MLGym-Bench, a new framework and benchmark for evaluating and developing LLM agents on AI research tasks. The key contributions of our work are: 🕹️ Enables the
16
120
494
@ClementDelangue
clem 🤗
9 months
I'd like to live in a world where all countries & organizations, big or small, can train their own AI, instead of just a few of the richest & biggest ones. This is IMO one of the most important topics ever and will shape the future of the world. That's why I'm super proud when
51
149
1K
@AlexGDimakis
Alex Dimakis
9 months
We are releasing OpenThinker-32B, the best 32B reasoning model with open data. We match or outperform Deepseek-R1-32B (a closed data model) in reasoning benchmarks. Congrats to Negin and the whole Open Thoughts team.
@NeginRaoof_
Negin Raoof
9 months
Announcing OpenThinker-32B: the best open-data reasoning model distilled from DeepSeek-R1. Our results show that large, carefully curated datasets with verified R1 annotations produce SoTA reasoning models. Our 32B model outperforms all 32B models including
8
25
210
@Aiims1742
Anirban Maitra
9 months
Full policy announcement here: https://t.co/McE4oyfu1d
@Aiims1742
Anirban Maitra
9 months
The NIH reduces its maximum indirect rate for research from 60% to 15% with immediate effect.
3
11
24
@praveersingh90
Praveer Singh
9 months
Wow this is big news !!!
@NIH
NIH
9 months
Last year, $9B of the $35B that the National Institutes of Health (NIH) granted for research was used for administrative overhead, what is known as “indirect costs.” Today, NIH lowered the maximum indirect cost rate research institutions can charge the government to 15%, above
0
0
2
@EricTopol
Eric Topol
9 months
Can your electrocardiogram be used to assess biological age? A new study suggests with A.I. it can be used to help predict age-related diseases, including not just cardiovascular, but also Alzheimer's and cancer https://t.co/6zTYXtEhTb open-access
16
115
370
@giffmana
Lucas Beyer (bl16)
9 months
I took a brief look at the Harmonic Loss paper tl;dr: instead of dot-product with softmax, do euclid dist with normalized 1/d**n. I kinda want this to work. I've dabbled with preferring euclid many times throughout my career (eg triplet loss etc) However...
@dbaek__
David D. Baek
9 months
1/9 🚨 New Paper Alert: Cross-Entropy Loss is NOT What You Need! 🚨 We introduce harmonic loss as alternative to the standard CE loss for training neural networks and LLMs! Harmonic loss achieves 🛠️significantly better interpretability, ⚡faster convergence, and ⏳less grokking!
27
147
2K
@Ar_Douillard
Arthur Douillard
10 months
We release today the next step for distributed training: --> Streaming DiLoCo with Overlapping Communication. TL;DR: train data-parallel across the world with low-bandwidth for the same performance: 400x less bits exchanged & huge latency tolerance
19
109
581
@praveersingh90
Praveer Singh
10 months
100% agreed. Also reinvigorate academic research in building resource contrained models !!!
@ClementDelangue
clem 🤗
10 months
False! Because it's open-source, fast and cheap, it will help create thousands of silicon valley startups, hundreds of thousands of jobs, massively step up competition and create hundreds of billions of value for silicon valley. IMO more positively impactful than any and all
0
0
1
@praveersingh90
Praveer Singh
10 months
Infact this can be managed by @NIH's Office of Science Policy. This will ensure zero publication costs to PI's, open access of public tax-funded research, regulated code & data sharing & most importantly happy reviewers (can also incentivize them through reviewer index metric).
@NiuSanford
Dr. Nina Niu Sanford
10 months
I think the NIH/NCI should create its own FREE open access journal and 1) Publish all NIH funded research 2) Pay reviewers ($ or some sort of credit system) Better quality reviews, research transparency (& less profit for publishing companies/predatory journals).
0
0
0
@madiator
Mahesh Sathiamoorthy
10 months
We are announcing Open Thoughts, our large-scale open-source effort to curate the best open reasoning datasets! DeepSeek-R1 is amazing but we still don't have access to high-quality open reasoning datasets. These datasets are crucial if you want to build your reasoning models!
46
291
2K
@nearcyan
near
10 months
cursor users pressing "accept all changes" until the bug goes away
250
773
11K
@AAOjournal
Ophthalmology
10 months
Validation of “EyeLiner” a deep learning pipeline to longitudinally evaluate and align color fundus photos, on 3 existing datasets. @AdvaithVeturi @kalpathy1 @praveersingh90 https://t.co/2fcQxBpfQZ #ophthalmology
0
6
8
@CUEyeCenter
CU Anschutz Department of Ophthalmology
10 months
Mary Labowsky, MD, assistant professor of ophthalmology, explains recent research that has shown that diabetic patients taking #Ozempic are at a higher risk of NAION and if patients should be worried. https://t.co/PmZpeFwjco
Tweet card summary image
news.cuanschutz.edu
Recent research shows that some diabetic patients have experienced NAION while taking weight loss drugs like Ozempic. Neuro-ophthalmologist Mary Labowsky, MD, explains the association and what...
0
1
3
@CUEyeCenter
CU Anschutz Department of Ophthalmology
11 months
Happy New Year from the Sue Anschutz-Rodgers Eye Center! Wishing you a prosperous 2025. Check out what we're looking forward to this year.
0
1
8
@CUEyeCenter
CU Anschutz Department of Ophthalmology
1 year
A new @CUEyeCenter research study has found that when Denver air quality is poor, visits to the Sue Anschutz-Rodgers Eye Center double. Authors @MalikKahook, MD, and Jennifer Patnaik, PhD, discuss the study at the link below. https://t.co/l9sJ1yRt64 #ClimateChange #Ophthalmology
Tweet card summary image
news.cuanschutz.edu
Ophthalmology researchers at the University of Colorado School of Medicine find that ocular surface irritation and allergy visits more than doubled along with higher concentrations of some pollutants.
0
2
7
@NIBIBgov
NIBIB
1 year
Tune in to the @MIDRC_ seminar on Oct. 15 to hear from @praveersingh90 about AI in medical imaging. Register here: https://t.co/HdJAE4xVge
0
1
1