Samra (Sam)

250 posts

Samra (Sam) banner
Samra (Sam)

Samra (Sam)

@Sam14_irshad

PhD candidate @ Kyung Hee University —Trying to find sweet spots.

south korea Katılım Nisan 2025
1.1K Takip Edilen74 Takipçiler
Samra (Sam)
Samra (Sam)@Sam14_irshad·
@SocialImpurity @mayukh_panja I see a lot of intersection between a game I used to play in primary school called ‘telephone game or Chinese whisper’ and how the conversations on Twitter evolve. Peoples tendency to mould someone’s response into a completely different take is mind-blowing.
English
0
0
0
466
Mayukh
Mayukh@mayukh_panja·
I don’t agree. A PhD student should not prioritize work-life balance. Getting to do a PhD is a privilege. You are paid to think. There is no pressure for you to be economically useful. It is a unique opportunity to push the boundaries of human knowledge and produce something ground breaking. And nothing great ever happens without complete devotion. Look at everything that moved and shaped the world. Every single person who created anything meaningful, in science, in arts, in music, in movies, devoted their lives to their craft. Extraordinary outcomes require extraordinary inputs and some degree of sacrifice. Sure, have work-life balance during your PhD. But be content a mediocre outcome.
Dr. Manabendra Saharia@m_saharia

Yesterday, I was giving an intro talk to our dept's new PhD students. Technical things aside, my number 1 suggestion has remained the same over the years: Treat your PhD like a job. - Avoid 1.5h lunch and three tea breaks. - Avoid gossiping and loitering at work. - Lab at 9 am and leave at 6 pm. Being productive till 11 pm in the lab is a lie people till themselves when their day starts at 1 PM. Everything worth doing can be done with high intensity focus during work hours. And having fun in life is the secret to being productive in a marathon.

English
359
277
3K
1.1M
Samra (Sam) retweetledi
Mathieu
Mathieu@miniapeur·
Mathieu tweet media
ZXX
1
6
57
1.5K
Samra (Sam) retweetledi
Shriram Krishnamurthi (primary: Bluesky)
Shriram Krishnamurthi (primary: Bluesky)@ShriramKMurthi·
Since I'm dunking on this, let me point to one insight here that's easy to miss. Don't sit around just to "look" productive. I found that I'm unproductive after lunch, but was dawdling in my office. Instead I turned 2-4 into my exercise time every day. Find yours, whenever it is!
Dr. Manabendra Saharia@m_saharia

Yesterday, I was giving an intro talk to our dept's new PhD students. Technical things aside, my number 1 suggestion has remained the same over the years: Treat your PhD like a job. - Avoid 1.5h lunch and three tea breaks. - Avoid gossiping and loitering at work. - Lab at 9 am and leave at 6 pm. Being productive till 11 pm in the lab is a lie people till themselves when their day starts at 1 PM. Everything worth doing can be done with high intensity focus during work hours. And having fun in life is the secret to being productive in a marathon.

English
3
6
164
17.5K
Samra (Sam)
Samra (Sam)@Sam14_irshad·
@miniapeur Even the people in academia don’t know anything about quality research, let alone the people in admin 🫠🫠🫠
English
0
0
0
20
Samra (Sam)
Samra (Sam)@Sam14_irshad·
@vineettiruvadi I agree phd is not like a job. However I think in original post, he is talking from the perspective of self-discipline.
English
1
0
13
812
Vineet Tiruvadi, MD PhD
Vineet Tiruvadi, MD PhD@vineettiruvadi·
Do not treat your PhD like a job - it isn't. It's a space for you to try to ask big questions, and that means lots of failing and learning to fail and bummer days and exploring different structures for your day.
Dr. Manabendra Saharia@m_saharia

Yesterday, I was giving an intro talk to our dept's new PhD students. Technical things aside, my number 1 suggestion has remained the same over the years: Treat your PhD like a job. - Avoid 1.5h lunch and three tea breaks. - Avoid gossiping and loitering at work. - Lab at 9 am and leave at 6 pm. Being productive till 11 pm in the lab is a lie people till themselves when their day starts at 1 PM. Everything worth doing can be done with high intensity focus during work hours. And having fun in life is the secret to being productive in a marathon.

English
4
12
251
21K
Fang Wu
Fang Wu@WUFang40615703·
Not all diffusion noise is equally useful for training!🤫 We introduce 🐯NoiseRater: a meta-learned framework that scores and selects informative noise instances during diffusion training. Instead of treating Gaussian noise uniformly, we learn which noise samples actually improve downstream generalization. Results: • Better FID on FFHQ + ImageNet • Improved training efficiency • Transfers across DiT backbone sizes A new axis for improving diffusion training: training-time noise valuation. Paper: arxiv.org/pdf/2605.08144 Great thanks to collaborators @Hanqun_CAO @XiangruTang @hcwww_ @MoleiTaoMath @KKuanPang @erranlli and my advisors @YejinChoinka @jure #AI #MachineLearning #DiffusionModels
Fang Wu tweet media
English
2
14
84
5.1K
Samra (Sam) retweetledi
kache
kache@yacineMTB·
you can outsource your thinking but you cannot outsource your understanding
English
256
3.8K
16.7K
2.4M
Samra (Sam) retweetledi
Jack Lu
Jack Lu@Jacklu_me·
“How’d neurips submission go?”
Jack Lu tweet media
English
0
1
57
2.9K
Siddhartha Gairola
Siddhartha Gairola@sidgairo18·
The void you feel after the deadline is over! 🥲 what do I do now ?
GIF
English
2
1
11
506
Tanmoy Mukherjee
Tanmoy Mukherjee@langer_han·
Good luck #NeurIPS authors. Hopefully everyone gets well deserved rest. Anyone has any recommendations on series to watch ?
English
7
0
39
6.2K
Samra (Sam) retweetledi
Ravid Shwartz Ziv
Ravid Shwartz Ziv@ziv_ravid·
My current NeurIPS submissions status...
Ravid Shwartz Ziv tweet media
English
15
5
211
31.9K
Samra (Sam)
Samra (Sam)@Sam14_irshad·
I can write better than Gemini, Claude, ChatGPT, and all of them together. 🐼
English
0
0
2
90
Samra (Sam)
Samra (Sam)@Sam14_irshad·
@miniapeur I may appear retarded to both (trained/untrained) eyes.
English
0
0
1
44
Samra (Sam)
Samra (Sam)@Sam14_irshad·
@oak_raj @gabriberton I won’t say it writes a better story but it definitely helps by providing alternative candidate phrases that I can use and improve the writing. But AI tools don’t really come up with better stories. I tell it ‘look into that direction and it looks’. That’s the best it does.
English
0
0
0
50
Rajvardhan Oak
Rajvardhan Oak@oak_raj·
@Sam14_irshad @gabriberton Yes, 100%. But also, an AI-written paper may be more impactful than the human-written one. Given an idea, the ‘story’ that it writes may be better than the one a human researcher writes.
English
1
0
0
60
Gabriele Berton
Gabriele Berton@gabriberton·
Again, this is bad for science, please don't do it
Amit LeVi@AmitLeViAI

Such a great evening to start a brand new research for NeurIPS in 3.5 days.🧘‍♂️ Day 1: planning. Night 1: running experiments and sending the abstract. Day 2: reading results fighting with Claude, and sending again. Night 2: sleep (optional). Day 3: opening Codex, and finally, write the pape in parallel. Night 3: resolving the “beef” with Claude (temporary peace) and going to sleep. Day 4: final reading, last-minute fixes, submission then some relaxation, maybe a beach walk. I’ll keep you posted on the results. This will be my only single-author paper, so I can’t hide behind other submissions if it gets rejected 😅

English
10
19
541
65.2K
Sarvesh Gharat
Sarvesh Gharat@SarveshGharat12·
@Arian_Khorasani Le me; Has been working since August on a paper, that will finally be submitted to NeurIPS PS; These conferences should seriously consider increasing their page limit though. Containing the stuff within 9 pages is so difficult
English
3
0
6
775
Arian Khorasani 🦅
Arian Khorasani 🦅@Arian_Khorasani·
That's definitely NOT the best & correct idea if you care about science, high-quality research work and being impactful for the community. Strong work takes time and must be taken seriously, and comes from a human's mind and not LLMs. I hope it never becomes normal.
Amit LeVi@AmitLeViAI

Such a great evening to start a brand new research for NeurIPS in 3.5 days.🧘‍♂️ Day 1: planning. Night 1: running experiments and sending the abstract. Day 2: reading results fighting with Claude, and sending again. Night 2: sleep (optional). Day 3: opening Codex, and finally, write the pape in parallel. Night 3: resolving the “beef” with Claude (temporary peace) and going to sleep. Day 4: final reading, last-minute fixes, submission then some relaxation, maybe a beach walk. I’ll keep you posted on the results. This will be my only single-author paper, so I can’t hide behind other submissions if it gets rejected 😅

English
2
1
70
8.3K
Samra (Sam)
Samra (Sam)@Sam14_irshad·
@stevenyfeng Honestly. This is my first Neurips submission and I spent 4 months on it. And still feeling shitty and uncertain about it. And I see this just 3 days before deadline.
English
1
0
23
1.8K
Steven Feng
Steven Feng@stevenyfeng·
This is the type of slop blowing up the number of submissions to conferences, screwing up the review process and wasting reviewers' time, and furthering the stochasticity of acceptances.
Amit LeVi@AmitLeViAI

Such a great evening to start a brand new research for NeurIPS in 3.5 days.🧘‍♂️ Day 1: planning. Night 1: running experiments and sending the abstract. Day 2: reading results fighting with Claude, and sending again. Night 2: sleep (optional). Day 3: opening Codex, and finally, write the pape in parallel. Night 3: resolving the “beef” with Claude (temporary peace) and going to sleep. Day 4: final reading, last-minute fixes, submission then some relaxation, maybe a beach walk. I’ll keep you posted on the results. This will be my only single-author paper, so I can’t hide behind other submissions if it gets rejected 😅

English
12
14
425
37.6K
Samra (Sam)
Samra (Sam)@Sam14_irshad·
@oak_raj @gabriberton The thing is novelty and impact does not solely come from results, it’s the way you have written your piece. Impactful writing takes time, unless your goal is to just get a paper published in a top-tier conference that no one would like to read or cite afterwards.
English
1
0
2
93
Rajvardhan Oak
Rajvardhan Oak@oak_raj·
@gabriberton Is it necessarily bad if the output is a valid research claim, tested on various benchmarks (at the same standards as any other paper) and has good results?
English
2
0
3
3.2K

Keşfet

@SocialImpurity @mayukh_panja @miniapeur @vineettiruvadi @Hanqun_CAO @XiangruTang @hcwww_ @MoleiTaoMath