Rishu Kumar

24.1K posts

Rishu Kumar banner
Rishu Kumar

Rishu Kumar

@rishdotuk

Long document understanding, Multilingual Evals and efficient models mainly, but other #NLProc applications in free time | vim enthusiast

Katılım Nisan 2011
1.4K Takip Edilen860 Takipçiler
Rishu Kumar
Rishu Kumar@rishdotuk·
@oliviawpy2023 I don’t mind people taking assistance from LLMs to construct a response. That’s good. But they should atleast spend time to see if the rebuttal actually makes sense. If there is “something obvious”™️ that the LLMs ignored from the comments.
English
1
0
3
222
Olivia Wang
Olivia Wang@oliviawpy2023·
@rishdotuk Unfortunately the thing is that there is no way we can ban this, but we can choose to do better ourselves.
English
1
0
1
238
Cigarette Nostalgia
Cigarette Nostalgia@CigsMake·
The deadly combination that creates empires and builds nations
Cigarette Nostalgia tweet media
English
263
2.8K
24.6K
1.1M
Rishu Kumar
Rishu Kumar@rishdotuk·
@riyaa189 If he was a friend, I’d have helped him figure out what’s going wrong.
English
0
0
0
153
Riya Agarwal
Riya Agarwal@riyaa189·
Whatever you do, don't be this guy. There's a person who every quarter messages me about how hard DSA is, how tough the market is, how he doesn't understand DP or graphs, consistently since the last 2 years. He wants to switch but won't put in any effort. Just messages.
Riya Agarwal tweet mediaRiya Agarwal tweet media
English
13
0
29
3.2K
Rishu Kumar
Rishu Kumar@rishdotuk·
@that_tallguy_1 That’s not true at all, it is a generic medicine store that’s for sure, but nothing of the sort that you are claiming/hyping it to be.
English
0
0
4
68
Priyanshu
Priyanshu@that_tallguy_1·
@rishdotuk Sir what I meant here is generic chemicals no other stuffs.
English
1
0
0
474
Priyanshu
Priyanshu@that_tallguy_1·
Got this all for just 850 from jan aushadi Kendra. They all contains pure products without any chemicals and fancy brandings.
Priyanshu tweet media
English
111
96
1.6K
190.9K
mish ⋆˙⟡
mish ⋆˙⟡@acidpapii·
are you guys “thik hai” or “thike” or “theek hai”
English
157
8
396
40K
Rishu Kumar
Rishu Kumar@rishdotuk·
@Arpitaaa01 If someone thinks trekking to a hilltop for photos is risky stuff, run like 2KM from them.
English
1
0
1
24
Arpita Das
Arpita Das@Arpitaaa01·
People hear “I like risky stuff” and think I mean trekking to a hilltop for photos. No bro, I mean flying a helicopter in another country and skydiving on a different continent. We are not talking about the same kind of adventure💀
English
7
0
37
1.4K
Coach Vibhav
Coach Vibhav@Coach_Vibhav·
My plate is free from guilt and tears.
Coach Vibhav tweet media
English
3
0
17
478
Sameera Pant
Sameera Pant@sameeraculous·
if you never get carsick, you are one of gods favourites for real. (every trek to / from blr airport really throws me off)
English
11
0
43
3.2K
lord.
lord.@dudexverse·
job se ko khush nhi h bc, jeevan ka asli maza kisi tier 4 city mein shaam ko 2 ghante walk krne mein hai.
Eesti
6
7
212
5.5K
Divyanshi
Divyanshi@divyanshifr·
chalo bhai aage ki strike koi aur continue karlo ek hi aadmi ko baitha diya hai itne din se
हिन्दी
3
0
3
300
Rishu Kumar retweetledi
Mads
Mads@europemaxxed·
Absolut genius
Mads tweet media
English
8
416
19.1K
197.9K
Rishu Kumar retweetledi
Tuhin Chakrabarty
Tuhin Chakrabarty@TuhinChakr·
A paper I am an AC on submitted 4 rebuttals in less than 10 mins from when the reviews came out. Students I am begging you. Please don’t do this. No body needs a skill.md to argue for a paper you wrote ( unless you delegated that to AI as well)
Lei Li@_TobiasLee

opensource our rebuttal skills, distilled from a lot of successful rebuttals from myself & labmates; also my experience serving as ACs for ARR. Hope this help for your EMNLP and incoming NeurIPS reviews :) github.com/TobiasLee/Rebu…

English
8
5
79
9.4K
Rishu Kumar
Rishu Kumar@rishdotuk·
Last exam “2.0” has a beautiful irony to it. :) Love the work though.
Dr. Datta M.D. (Radiology) ✈️ Switzerland @AI4Good@DrDatta_AIIMS

🔥Today, we are releasing one of the first visual reasoning benchmarks for autonomous AI diagnosis in healthcare! 🚀Introducing Radiology’s Last Exam 2.0 (RadLE 2.0) from @CRASHLabAI, an uncertainty-aware benchmark for autonomous diagnosis in radiology! ✅In the last few days, the AI frontier has moved significantly. @OpenAI launched GPT-5.6 Sol. @Meta launched Muse Spark 1.1. @xAI dropped Grok 4.5. 🙌We’ve benchmarked all frontier, open-source and medical VLMs in RadLE2.0 and the leaderboard is now LIVE! 🚨 Before AI models are handed autonomy, one question matters more than any accuracy score: Do they know when to STOP and hand over to a human? ⚠️ A confident wrong diagnosis is far more dangerous than an honest “I don’t know.” Yet most models are bad at admitting the latter! 🚀 We release five RadLE 2.0 Scores: Confidence Weighted, Reliability, Accuracy, Safety and Handover Readiness and we find that models from @OpenAI @AnthropicAI @MetaAI @GoogleDeepMind @xAI @nvidia @Alibaba_Qwen @MistralAI @MiniMax_AI all score very differently as they optimize for different metrics! 🚨But most importantly, NONE of the Models have been able to reach the average human expert baseline! ⚡️A thread on what we found and which models aced our metrics! Link to the leaderboard and technical report at the end of the thread!

English
0
0
4
916
Phalgun
Phalgun@phalgooon·
an angel loses a wing anytime you send a docx to someone over email
English
4
0
59
4K

Keşfet

@oliviawpy2023 @CigsMake @riyaa189 @that_tallguy_1 @acidpapii @Arpitaaa01 @Coach_Vibhav @sameeraculous