Aditi Jha (@aditi_jh) 's Twitter Profile
Aditi Jha

@aditi_jh

Computational neuroscientist | Postdoc @StanfordBrain | Formerly PhD @Princeton with @jpillowtime

ID: 4537696110

linkhttp://aditijha7.com calendar_today12-12-2015 12:27:15

184 Tweet

730 Takipçi

496 Takip Edilen

Scott Linderman (@scott_linderman) 's Twitter Profile Photo

Did you know that you can parallelize *nonlinear* RNNs over their sequence length!? Our NeurIPS Conference paper "Towards Scalable and Stable Parallelization of nonlinear RNNs," which introduces quasi-DEER and ELK to parallelize ever larger and richer dynamical systems! 🧵 [1/11]

Did you know that you can parallelize *nonlinear* RNNs over their sequence length!? Our <a href="/NeurIPSConf/">NeurIPS Conference</a> paper "Towards Scalable and Stable Parallelization of nonlinear RNNs," which introduces quasi-DEER and ELK to parallelize ever larger and richer dynamical systems! 🧵 [1/11]
Dan Biderman (@dan_biderman) 's Twitter Profile Photo

How can we use small LLMs to shift more AI workloads onto our laptops and phones? In our paper and open-source code, we pair on-device LLMs (ollama) with frontier LLMs in the cloud (@openai, @together), to solve token-intensive workloads on your 💻 at 17.5% of the cloud cost

Andrew Drozdov (@mrdrozdov) 's Twitter Profile Photo

We're probably a little too obsessed with zero-shot retrieval. If you have documents (you do), then you can generate synthetic data, and finetune your embedding. Blog post lead by Jacob Portes shows how well this works in practice. databricks.com/blog/improving…

Annie G. Bryant (@anniegbryant) 's Twitter Profile Photo

Excited to share work co-led with Aditi Jha developing a data-driven selection technique for overlapping community detection algorithms, applied to the human structural connectome! biorxiv.org/content/10.110…

Excited to share work co-led with <a href="/aditi_jh/">Aditi Jha</a>  developing a data-driven selection technique for overlapping community detection algorithms, applied to the human structural connectome!

biorxiv.org/content/10.110…
Scott Linderman (@scott_linderman) 's Twitter Profile Photo

Excited to share our JOSS paper (joss.theoj.org/papers/10.2110…) and code (github.com/probml/dynamax)! Special thanks to our reviewers — Dynamax v1.0 is much improved thanks to their feedback!