@LucaAmb @miniapeur So does getting an email approval from all authors during the upload process solve that?
English
Sahaj Singh Maini
67 posts
Sahaj Singh Maini
@SahajSinghMaini
Computer Science x Cognitive Science PhD candidate at Indiana University-Bloomington @IUBloomington @IULuddy He/Him/His.
Bloomington, IN Katılım Eylül 2019
1.1K Takip Edilen81 Takipçiler
Note that, after being banned, you risk to be shunned by your collaborators as you would block them from sharing on arxiv otherwise
Also your students will be punished as well
All of this because one of your sleep deprived grad student sent an arrive update without even asking you
English
Although I generally feel something should be done about arXiv, I can understand the backlash given the severity of the proposed punishment. Could we at least institute a points-based system, like for a driving licence? A direct one-year ban, plus losing the ability to share unpublished work on arXiv, seems too harsh.
That said, failing to notice that a reference is incorrect is different from submitting a paper that contains an AI prompt, which could literally be detected by reading the said paper.
Christopher D. Long 🇺🇦🏳️🌈🌹@octonion
The backlash against arXiv is a bit odd. All they're asking is that you read your papers before submitting them.
English
are good supervisors more important than the phd topic or are they equally important
English
@RylanSchaeffer But would u prefer a MS or PhD student who is a masterful Opus 4.6 wrangler 🤔?
English
I highly respect Noam but I strongly disagree with his sentence "I don't think Opus 4.6 is actually at the level of a remote entry-level AI researcher"
In a non-hypothetical head to head comparison, I prefer Opus 4.6 over Stanford CS MS & most PhD students
Noam Brown@polynoamial
I appreciate @Anthropic's honesty in their latest system card, but the content of it does not give me confidence that the company will act responsibly with deployment of advanced AI models: -They primarily relied on an internal survey to determine whether Opus 4.6 crossed their autonomous AI R&D-4 threshold (and would thus require stronger safeguards to release under their Responsible Scaling Policy). This wasn't even an external survey of an impartial 3rd party, but rather a survey of Anthropic employees. -When 5/16 internal survey respondents initially gave an assessment that suggested stronger safeguards might be needed for model release, Anthropic followed up with those employees specifically and asked them to "clarify their views." They do not mention any similar follow-up for the other 11/16 respondents. There is no discussion in the system card of how this may create bias in the survey results. -Their reason for relying on surveys is that their existing AI R&D evals are saturated. Some might argue that AI progress has been so fast that it's understandable they don't have more advanced quantitative evaluations yet, but we can and should hold AI labs to a high bar. Also, other labs do have advanced AI R&D evals that aren't saturated. For example, OpenAI has the OPQA benchmark which measures AI models' ability to solve real internal problems that OpenAI research teams encountered and that took the team more than a day to solve. I don't think Opus 4.6 is actually at the level of a remote entry-level AI researcher, and I don't think it's dangerous to release. But the point of a Responsible Scaling Policy is to build institutional muscle and good habits before things do become serious. Internal surveys, especially as Anthropic has administered them, are not a responsible substitute for quantitative evaluations.
English
Sahaj Singh Maini retweetledi
Vision-language models learn the geometry of human perceptual space. arxiv.org/abs/2510.20859
English
Sahaj Singh Maini retweetledi
We need new rules for publishing AI-generated research. The teams developing automated AI scientists have customarily submitted their papers to standard refereed venues (journals and conferences) and to arXiv. Often, acceptance has been treated as the dependent variable. 1/
English
Sahaj Singh Maini retweetledi
Sahaj Singh Maini retweetledi
I am currently holding my dad's cryopreserved brain tumor samples in hopes of creating a personalized vaccine for immunotherapy. However, there are some critical and time-sensitive questions in the attached post: x.com/tejasdkulkarni…
This is time-sensitive so would appreciate any DMs/RTs.
Tejas Kulkarni@tejasdkulkarni
English
@YiTayML @kohjingyu I have personally found tensorboard slow to load previous runs. I sync my tensorboard and wandb. Best of both worlds.
English
@kohjingyu when i first left G i saw wandb in some default code and i was like "wtf is that" and had this knee jerk reaction of trying to remove it just like how you would remove a fly from soup.
no one should use anything other than tensorboard.
English
Do people really like WandB, or are you all just messing with me?
I've used it for just a week, but the UX seems so much worse than Tensorboard. I really like the automatic GPU utilization logs though.
English
Sahaj Singh Maini retweetledi
so many mainstreamers pumping AI now.
a good filter is to check if they were pumping crypto before lol.
English
@YiTayML haha, I know. it was a perfect reply to "Offend a ML Researcher in one tweet."
English
We train Transformers to encode algorithms in their weights, such as sorting, counting, and balancing parentheses from lots of data.
I never thought we may also go in the *reverse* direction: *compile* Transformer weights directly from explicit code! Cool paper @DeepMind:
1/🧵
English
Sahaj Singh Maini retweetledi
Success is failing in a way that leads clearly to if-then learnings.
English
Sahaj Singh Maini retweetledi
Did you know @Meta has transitioned @PyTorch to the Linux Foundation?
Read Mark Zuckerberg's post about the launch of the PyTorch Foundation here: hubs.la/Q01m2X8c0
#PyTorchFoundation #PyTorch #AI #OpenSource
English
Sahaj Singh Maini retweetledi
"Progress in science depends on new techniques, new discoveries and new ideas, probably in that order"
--- Sydney Brenner (~1980)
English
English
I have a weirdly strong memory of someone sharing a very very large RL gym environment, which looked a lot like NES Zelda, like +/- 9 months ago. Does anyone have anyone have any idea what I’m talking about? 😅
English
Sahaj Singh Maini retweetledi
This optical illusion – one of the best of the last few years – will really mess with your head. How many black dots are there? The answer is 12, but your visual system won't let you see them all at once. Explanation here: theverge.com/2016/9/12/1288…
English
@DrSJaishankar, not choosing a side in such desperate times is equivalent to siding with the aggressors. I vehemently condemn this decision like many other Indians who share my values. Such a decision places our country on the wrong side of history. #StandWithUkraine
English