Nishant Mehta

Prof of CS at @uvic, machine learning theory research person

[Nishant Mehta @mehtalearner]

@WangMa70190365 Here's a general rule: if fixing the concerns requires a major revision of the paper, it's like most appropriate to revise and resubmit elsewhere. You're quite lucky if the reviewers/AC are ok with a major revision. I've been a lucky author before, but it indeed felt quite lucky!

[Nishant Mehta @mehtalearner]

@CitImmCanada What actions is IRCC taking to solve the problem of seemingly indefinite delays for security checks?

[Nishant Mehta @mehtalearner]

@peterjansen_ai Idea: have an oral exam for accepted papers (exam at the conference). If authors don't pass the oral exam, the paper is auto-retracted. The beauty is that the shame (of claiming to author a paper without understanding the ideas) is public b/c the paper was initially accepted.

[Nishant Mehta @mehtalearner]

@yoavgo @_michaelginn for sure I can easily identify bigger issues (reviewers offloading to LLMs, for example; even worse, students <-- next generation of reviewers offloading their education to LLMs)

[Nishant Mehta @mehtalearner]

@bremen79 @NicoloCB Thanks prof : - )

[Nishant Mehta @mehtalearner]

@she_ide @valdombre @DominionReview I didn't say Canada owes anyone anything. Just by having an opinion you claim I am (or the opinion is) entitled... Also, I'm not sure if discussing privilege is constructive; who cares about that? Finally, brilliant people learning from each other is good; why isolate Canadians?

[Nishant Mehta @mehtalearner]

@arjunswritings @USAmbCanada If this is satire, you're doing well.

[Nishant Mehta @mehtalearner]

@valdombre @DominionReview Another great benefit: stopping many top international students of the opportunity to come to Canada. The beauty of the harsh restrictions is the far fewer talented students feeding their ingenuity into Canadian research and industry. Bonus: profs and universities are struggling.

[Csaba Szepesvari @CsabaSzepesvari]

Very happy to see a familiar name here. Congrats Gabor Lugosi who knows better not to be on this platform;)

[Nishant Mehta @mehtalearner]

@lpachter @WKCosmo Meta-hallucination: when someone hallucinates that someone else hallucinated a reference. The meta-conversation now shifts to meta-banning (or is discussion of meta-banning meta-meta-banned?)

[Nishant Mehta @mehtalearner]

@CHaesemeyer @VinceVatter Exactly! Rather than "errors", we're talking about authors that are using broken processes. I find it really satisfying to incentivize someone toward a better process (one where you read your own paper)

[Andreas Kirsch 🇺🇦 @BlackHC]

In a personal capacity, and not on behalf of Google or Google DeepMind: very glad I could speak with FAZ and SZ about Google's reported classified AI deal with the Pentagon.

Autonomous weapons, mass surveillance, and autonomous policing all concern public health and safety. They are matters of enormous public interest, not just internal policy debates.

Companies that should know better are using safety-washed PR that undermines the honest public conversation we need about AI in classified settings.

[Nishant Mehta @mehtalearner]

@LucaAmb @thomasfbloom Is this a joke? --> "copying pasting output is the correct way to use an LLM for editing"

Another analogy: I allow you to dance on stage in front of an audience, but if you slip and fall because you don’t know what you’re doing, then I won’t let you into my auditorium again.

[Nishant Mehta @mehtalearner]

@taiyasaki Indeed bullshit. That said, technically, you can just take a photo of the poster, make that a frame, and repeat the frame for a 5 minute silent video. A literal video presentation of the poster...

[Nishant Mehta @mehtalearner]

@skye7821 @AmitLeViAI What better sign of excitement and confidence in a submission when in the face of severe time scarcity one posts on X instead of working on their submission. The beauty is that more engagement with this guy on X reduces the chance of his paper getting in.

[Nishant Mehta @mehtalearner]

@AmitLeViAI @icmlconf Wow, but I'd be curious to see the meta-review text.

[Nishant Mehta @mehtalearner]

@demisama_ A priori of the decision text, it definitely looks bad! Yet, what was the text of the decision? Without that, it's hard to know if it was a bad call or not. A lot of reviewers are submitting shit reviews which makes a scrupulous AC effectively take over for the real evaluation

[Nishant Mehta @mehtalearner]

@Tigger0000 @Emil126256458 @NeurIPSConf By the nature of my skill set, I may be a de facto member of the labor aristocracy. Yet, I do support those who are not. My preference would be for anyone (regardless of income or nationality) to be able to learn without limits and participate in research. Much respect

[Nicolò Cesa-Bianchi @NicoloCB]

We’re thrilled to announce the:

🧠 NExT-Game: New Frontiers of Game-Theoretic Learning📍Seoul | 🗓 July 2026 | 🤖 Worskshop@ICML

👉https://t.co/MMZv5kgzfg
📝 Submission portal: OpenReview
📅 Submission Deadline: April 24, 2026 (AOE)
📧 [email protected]