PhD student at @HuthLab. Interested in language, computers, and the brain. No thoughts about memes.
Joined June 2021
- Tweets11
- Following735
- Followers269
- Likes61.9K
Aditya Vaidya retweeted
Excited to present our work on bilingual brain-informed fine-tuning at #NeurIPS2025 🎉
neurips.cc/virtual/2025/loc/…
We fine-tune monolingual and multilingual language models using fMRI data from bilingual participants as they read naturalistic stories in English and Chinese.
Aditya Vaidya retweeted
We’re excited about our first paper looking at speech encoding in single neurons across the depth of human cortex. Out today in @nature! nature.com/articles/s41586-0… [1/6]
If you're at CoNLL / EMNLP check out our poster in the afternoon!
Code and data are also now available here: github.com/HuthLab/lm-repeat…
At CoNLL poster session 1 today: @_avaidya arxiv.org/abs/2310.06408 LM and human next-word predictions are correlated, but not when text repeats! This points to divergence btwn memory mechanisms, which we localize in the model and then modify to get more human-like behavior
If you're still around at #SfN2023, come by my poster in the afternoon!
At the final SfN poster session @_avaidya will present some new work he’s doing w/ me & @LibertySays. It sounds a little nuts, but we’re using in silico experiments on fMRI models to replicate effects that really should only work in ECoG. Poster DD15 Wed PM.
Aditya Vaidya retweeted
Our language decoding paper (@AmandaLeBel3 @shaileeejain @alex_ander) is out! We found that it is possible to use functional MRI scans to predict the words that a user was hearing or imagining when the scans were collected nature.com/articles/s41593-0…
Aditya Vaidya retweeted
Excited to present a comprehensive analysis on the extent to which deep neural network (DNNs) can account for brain responses in the human auditory cortex! Work with @jenellefeather (co-1st) @dlboebinger @JoshHMcDermott
biorxiv.org/content/10.1101/…
*Today* at ICML, I'll be presenting our work on predicting brain data with self-supervised speech models. Drop by the poster or spotlight talk if you're interested, or just want to chat!
New paper from @_avaidya & @shaileeejain just accepted at ICML: Self-supervised models of audio effectively explain human cortical responses to speech
arxiv.org/abs/2205.14252
Aditya Vaidya retweeted
If you’ve ever tried to accurately reconstruct hand pose while someone is holding an object from RGBD video, you know it’s kinda bad, right? Well, we (@akarshkumar0101 & @_avaidya) have make it work a bit better! arxiv.org/abs/2105.08196 (accepted @ CVPR EPIC Workshop)