Sabitlenmiş Tweet
Every (almost) talk in AI conferences these days: we prompt a LLM in a very specific (and weird) way to solve a task and achieve SOTA performance (ofcourse there is also some theory behind it). #ijcai2024
English
Mayank Nagda
146 posts
Mayank Nagda
@MayankNagda35
Trying Gambits. Doing Blunders.
Kaiserslautern, Germany Katılım Temmuz 2013
77 Takip Edilen52 Takipçiler
Mayank Nagda retweetledi
Breaking News: #NeurIPS2025 received over 25,000 abstracts. Top submitters include:
ChatGPT Univ: 6,500
Claude Inst of Tech: 4,255
Gemini Polytechnic: 3,135
DeepSeek R1 labs: 2,300
Llama Herd: 1,700
These groups also graciously agreed to do most of the reviewing..
English
@vvenki22 @abk_tau @rikvannoord Very stupid. They asked me to do significance tests. I have already done so and it is mentioned as strength by 2 reviewers. Other comments are about improving the limitations section.
English
@MayankNagda35 @abk_tau @rikvannoord How are the comments in meta review ? i.e suggested revisions section
English
We submitted a paper to ARR ACL. Got a meta review of 4, but the paper was still rejected. Fine, it happens. We improve the paper based on the comments, resubmit to EMNLP, and ask for same reviewers and AC. Reviews are the same, great.
English
@abk_tau @rikvannoord I have same scores for NAACL (4,3.5,3) and meta-review of 3. What happened with your paper?
English
@rikvannoord Also happened to us in the last cycle.
All reviews are positive (4,3.5,3), concerns were addressed in the rebuttal (+ acknowledged by the reviewers!), yet the meta-review completely ignores the rebuttal, contradicts itself and gives us a 3 (=major revision needed).
Frustrating.
English
@Michael_J_Black @CVPR Thank you for making this public. I did not even know one could submit a video. Does it not breach anonymity policies in some way (with regard to the author’s voice)? It feels like including a video is a hack that very few people are using.
English
Many people are in the middle of the @CVPR deadline. So I'm sharing my guide to writing a CVPR paper (or any paper). My students have had this for years but I haven't shared it publicly before. I hope you find it useful and write a great paper. #CVPR2025 @black_51980/writing-a-good-scientific-paper-c0f8af480c91" target="_blank" rel="nofollow noopener">medium.com/@black_51980/w…
English
Attending a topic modeling talk by David Blei at #ijcai2024 with a topic modeling paper (poster) in hand.
English
It makes sense why Reviewer 2 keeps asking me about solving a task I am working on by prompting a LLM.
English
Every (almost) talk in AI conferences these days: we prompt a LLM in a very specific (and weird) way to solve a task and achieve SOTA performance (ofcourse there is also some theory behind it). #ijcai2024
English
Excited to present our paper on integrating syntax with neural topic models at @IJCAIconf 2024 in Jeju Island, South Korea! If you're around, join me for the talk and poster session on August 7th at 15:00 in the NLP session. #ijcai2024
Jeju-do, Republic of Korea 🇰🇷 English
🎊 Happy to announce that my work is accepted at @IJCAIconf.
ML@RPTU@mlrptu
🎉 Excited to announce our two papers accepted at IJCAI 2024 (@IJCAIconf): The papers tackle crucial challenges in language understanding and multimodal learning. We will be presenting them at the conference. #IJCAI2024 #AI #NLP #MultimodalLearning
English
@code Could you please refrain from sending out updates to your app every other day? It's quite annoying.
English
Mayank Nagda retweetledi
I think I should be allowed to cage fight my EMNLP reviewers instead of having a rebuttal period tbh
English
checkout our latest paper: arxiv.org/abs/2306.00539
ML@RPTU@mlrptu
📢 New paper alert! Our latest research selected at #ACL2023 findings, calls for standardization and validation in Text Style Transfer evaluation. We've conducted a comprehensive meta-analysis, revealing notable gaps in both human and automated evaluation processes. #NLProc (1/n)
English
Reducing fonts in the search bar while scrolling down is bad UI @Google
English
@baohao_liao @MayankNagda35 I recommend not trying to hack the link. They track these activities and you may get banned. I got a warning once.
English
@prajdabre1 Or maybe not all tracks are updated (:fingers crossed for my other paper:).
English
@MayankNagda35 Congratulations! By that logic all my papers are rejected! 🥲
English