PhD student at @EdinburghNLP | Previously @ The Boston Dynamics AI Institute @rai_inst, @nyuniversity, @AmazonScience, @CERN, @iiit_hyderabad
Edinburgh, Scotland
Joined March 2020
- Tweets48
- Following1.6K
- Followers124
- Likes363
Pinned Tweet
Excited to share that I will be presenting my 1st year PhD work at #ICLR2026 @iclr_conf next week in Rio 🇧🇷
❓In-Context Learning (ICL) has been a widely used paradigm to adapt LLMs for a task with a few examples or shots. But does it also work for multimodal examples? 🧵
Akash Gupta retweeted
Excited to share my recent work @ServiceNowRSRCH ! We introduce a new privacy-centric deep research dataset and show models frequently leak enterprise information.
However, training with dense _situational_ rewards efficiently learns to jointly optimize performance and privacy
Happening now!! Drop by if you are interested in test-time learning or making models perform better with limited data!!
📍Pavilion 3 , P3 - #206
@iclr_conf #ICLR2026 @EdinburghNLP
Excited to share that I will be presenting my 1st year PhD work at #ICLR2026 @iclr_conf next week in Rio 🇧🇷
❓In-Context Learning (ICL) has been a widely used paradigm to adapt LLMs for a task with a few examples or shots. But does it also work for multimodal examples? 🧵
If you would like to know more, check out the links below:
📜 Paper: arxiv.org/abs/2506.06905
💻Github: github.com/akashgupta97/MAPD
Or stop by our poster on the first day -
🕞 Thurs, 23rd April 2026 - 3:15 PM - 5:45 pm -03
📍Location: Pavilion 3 P3-#206 at the Riocentro Convention
Akash Gupta retweeted
Do you remember a few years back when the notion of *Inductive Bias* was central to machine learning and AI research?
I just wrote: *The Bittersweet Lesson* 😁 The strange case of inductive bias in Transformers
docs.google.com/document/d/1…
Comments/feedback welcome!
Akash Gupta retweeted
Today, I’m excited to introduce @DeepMind's AlphaDev, published in @Nature. It discovered faster sorting and hashing algorithms which have been released in standard open source libraries. But how does AlphaDev works? 🧵
dpmd.ai/3MMdUmE
#ICASSP2023 @ieeeICASSP Excited to announce that I will be presenting my work at @AmazonScience on modality-preserving embedding for audio-video synchronization at ICASSP 2023!! Do check out our poster if you're attending!
Paper: ieeexplore.ieee.org/abstract…
Akash Gupta retweeted
Christopher Nolan doesn’t use detailed outlines.
Instead, he “draws shapes and diagrams and other structural things.”
Like this — the plot map Nolan used for Inception.
Here’s a breakdown:
Akash Gupta retweeted
Story time: i had a prompt engineering battle against @karinanguyen 🧵
Akash Gupta retweeted
#AlphaCode: a drastic version of seq2seq & search. The Bitter Lesson strikes again, exposing the power of general methods that scale w/ increased compute! Like AlphaStar, an agent outperforms me at something I love, competitive programming 🤣 @codeforces
dpmd.ai/Alpha-Code
Akash Gupta retweeted
Combining a traditional method & deep learning for binding pocket prediction
@AggarwalRishal @aksgupta97 @IHUB_Data @iiit_hyderabad #ml4science #MachineLearning #compchem #drugdiscovery
DeepPocket: #Ligand_Binding_Site Detection and Segmentation using 3D Convolutional #NeuralNetworks
pubs.acs.org/doi/10.1021/acs…
@AggarwalRishal @deva_priyakumar
#ASAP
Akash Gupta retweeted
DeepPocket: Ligand Binding Site Detection and Segmentation using 3D Convolutional Neural Networks pubs.acs.org/doi/10.1021/acs…
@AggarwalRishal @aksgupta97 @iiit_hyderabad @IHUB_Data @JCIM_ACS #compchem #ml4science #drugdesign @machinelearning