Gorsha Sur

Lawyer, Athlete, Techie, Good People

[Kirill Igumenshchev @kirill_igum]

"Does Synthetic Data Generation of LLMs Help Clinical Text Mining?" is a good step towards practical applications of #LLM to #dataprivacy fields like #health and #law. Given a few examples, LLM can generate a synthetic dataset that significantly improves the LLM performance. The paper is about the health field, but its focus on name entity recognition and relation extraction is highly applicable to the #lawdiscovery process.
https://t.co/nETL4wvA37
I'm looking forward to applying this insight.

[Cartesi @cartesiproject]

5/ And finally, the biggest congratulations goes to our first-place winner @kirill_igum with his project Teach AI! What is it? Lucky for us, he explained his DApp in a quick little video: https://t.co/RkVHfycvJL

[CodeX Stanford @CodeXStanford]

Codex is happy to co-sponsor the @Stanford AI Symposium--apply to join us today!

[Kirill Igumenshchev @kirill_igum]

me and @GorshaSur presenting our #LLM #LAW project

[CodeX Stanford @CodeXStanford]

Join us at our next CodeX Group Meeting happening TOMORROW (June 8) at 1:30pm featuring our guests,
@AnthonyNovaes_ and @kirill_igum & Gorsha Sur. For more information, see: https://t.co/FTuUQ914I9.

[Gorsha Sur @GorshaSur]

Fun times participating in last week’s LLMxLaw hackathon organized by @CodeXStanford and winning Best NonTechnical with Total Trial Prep project developed together with data guru @kirill_igum to facilitate complex litigation preparation. Use case is inspired by @coe Register.

[CodeX Stanford @CodeXStanford]

Congratulations to Total Trial Prep for winning Best NonTechnical

[Gorsha Sur @GorshaSur]

I'm reading: Stanford CodeX Hosts the First LLM x Law Hackathon at https://t.co/0f9318iGte via @StanfordLaw

[CodeX Stanford @CodeXStanford]

Our LLM x Law Hackathon this weekend brought together some the brightest and most innovative minds to think creatively and collaboratively on how we could integrate LLMs in the legal space.

[Kirill Igumenshchev @kirill_igum]

#CodeXStanford hosted an amazing #LAWxLLMhackathon.
@GorshaSur and I won. Our Project:
Total Trial Prep creates a curated pool of supporting witnesses, evidence, and visual aids from unorganized data using #LLM and human feedback. Our use case created a Register of Damages, a list of evidence, fact and expert witnesses, a complaint, a simulated response, and trial visual aids as related to the Russia-Ukraine war and based on the rules of #InternationalCourtOfJustice. A large amount of repetitive work is a perfect application for AI. Without AI, junior-level legal staff would process 100K+ text and image evidence costing $100M+ in legal fees. This use case can be adapted for complex #litigation, such as #classaction, product liability, or human rights violation, to identify prospective claimants, witnesses, and evidence and assist in trial preparation.

[Gorsha Sur @GorshaSur]

This is an excellent article by @LoisElfman about me joining forces with @jeffbenz and Jennifer Yuen to launch a law…https://t.co/BcJbBIC4sE