Saeed Shurrab

While alive:
Keep learning

[Malikeh Ehghaghi @Malikeh5]

In a field moving so fast that ideas get outdated days after release, we are wasting enormous time and energy on a closed-door conference lottery. Just as companies are moving away from traditional coding interviews, research needs a revolution in how work is presented, reviewed, and shared. Traditional conferences no longer work.
#NeurIPS2026 #ICML2026 #ICLR2026

[Saeed Shurrab @saeedsh91]

If reviewers did not engage during Phase 2, will they engage during Phase 3?
And if they already engage during Phase 3 with the Area Chair, it would be more reasonable that they got engaged during Phase 2, while we are able to respond.
#NeurIPS @NeurIPSConf

[Saeed Shurrab @saeedsh91]

@pratzohol @NeurIPSConf Yeah, the same for me as well, no response with one reviewer changed the his/her review substantially on 31st July after adressing his comments

[Saeed Shurrab @saeedsh91]

@prathoshap In my case, one of the reviewers has substantially changed his review narrative, after posting my rebuttal, by removing the strength and providing a lengthy critique as the paper submitted as use-inspired but does not satisfy the requirements of this category

[prathosh A P @prathoshap]

Neurips Author/reviewer discussion phase is going on. Both in my papers and my AC stack, there's hardly any discussion going on! The community has to wake up - a submission at these venues are months of hard work of several people - they deserve a response at the least!

[Danish Pruthi @danish037]

This year at @iclr_conf, we've policies where reviewers would be required to (a) disclose AI tools used for assisting in writing peer reviews, and (b) report original self-written assessment and any LLM interactions.

[Saeed Shurrab @saeedsh91]

@abursuc Still not a not a solution, borderline papers that do not get substantial improvement in score can still risk it withdraw and submit to ICLR

[Saeed Shurrab @saeedsh91]

Is it really fair that reviewer ask for additional baselines after comparing 15+ baselines of varied strength? The setup published relevant work is not straightforward almost always.
#NeurIPS2026

[Zhaoning Wang @Zhaoning_Eric_W]

I honestly believe that large AI Venues like NeurIPS should have a credit system. People get lower credit when they start "trashing" the submission, forcing people to take each submission seriously

[Saeed Shurrab @saeedsh91]

@yisongyue @NeurIPSConf I posted 10 k rebuttal

[Saeed Shurrab @saeedsh91]

@JeanKossaifi Super!

[Jean Kossaifi @JeanKossaifi]

PSA for anyone preparing a NeurIPS rebuttal (or any OpenReview response).

Draft in Google Docs, go to Tools > Preferences > Enable Markdown.
When you’re done you can simply select the text, right-click, and “Copy as Markdown”.

No more fixing formatting by hand. You’re welcome!

[Saeed Shurrab @saeedsh91]

@sumeetrm I got almost the same ratio in my reviews

[Saeed Shurrab @saeedsh91]

@HaritzPuerto Rebuttal VS weekend!
Rebuttal wins hahahaha

[Saeed Shurrab @saeedsh91]

@Arnobg32 Does any exceeds 4 by first submission?

[Saeed Shurrab @saeedsh91]

TRUE!

[Chhavi Yadav @chhaviyadav_]

So the per-review rebuttal buttons are only available until 27th AoE?? If true, this is really bad on multiple levels.

Firstly, only 3-4 days to do experiments that reviewers asked for? That’s very little time! — what’s even the point of giving this time when experiments can’t finish? Reviewers are mostly not satisfied by words, they need results!!

Secondly, the first day the reviews were released late , some people weren’t able to see them for one whole day! I could see the meta-reviews only the next day..

Most importantly, why are we expecting everyone to work on the weekends??? Students already have bad work-life balance and this is not helping! @NeurIPSConf

[Saeed Shurrab @saeedsh91]

Our latest work on retrieval-augmented foundation models for EHR data:
EHR-RAGp: Retrieval-Augmented Prototype-Guided Foundation Model for Electronic Health Records
https://t.co/AapEdadcfy
#AI #HealthcareAI #MachineLearning #DeepLearning #EHR #ClinicalAI #FoundationModels #RAG

[Saeed Shurrab @saeedsh91]

#NewProfilePic

[Clinical AI Lab @NYUClinicalAI]

#Meet_the_team! @abudaoudmouath is a Bioengineering Undergraduate working on the development of an Arabic telehealth LLM. He is an aspiring physician with a passion for integrating AI into medicine to increase access to healthcare. Connect to discuss the future of AI in medicine!

[Saeed Shurrab @saeedsh91]

@abdduidti