Hey #CausalTwitter, some news about our DoWhy python library - we are excited by how people are using DoWhy and we are working to support a broader community!
To broaden access to this important causal library—widely used with 1M downloads—DoWhy is moving to an independent PyWhy GitHub organization. The first step: collaboration with @awscloud, which is contributing advanced graphical causal model algorithms: msft.it/6010bVMDm
✨I’m on the academic job market!✨
I'm a CS Postdoc at Stanford in the @stanfordnlp. I develop AI and causal inference to tackle societal issues like online civility, racial justice, sustainability, and more.
Check out the thread for details!
Website: kristinagligoric.com/
What happens when LLMs encounter information that contradicts their static knowledge? 🤔
Discover our findings, including a new dataset and interpretability method, in our ACL 2024 paper! 🧵👇
📄 Read the paper: arxiv.org/abs/2312.02073
🖥️ Explore more: epfl-dlab.github.io/llm-grou…
Check out our new work, now out in Frontiers in Nutrition:
"Measuring and Shaping the Nutritional Environment via Food Sales Logs: Case Studies of Campus-Wide Food Choice and a Call to Action"! frontiersin.org/articles/10.…
We often wonder how the awareness of being "watched" affects our online behaviors. Our #CHI2024 paper studies observer (/hawthorne) effect: how do people's social media use deviate from typical use with the awareness of being observed? w/ Pranshu @emrek@GloriaMark_PhD@munmun10
Excited to release the AI Controller Interface, making it easier to tightly integrate and experiment with new strategies for controlling #LLM generation with custom logic, data, constrained decoding, parallel generations, ... cc: @m_moskal
The AI Controller Interface helps researchers and developers to efficiently implement existing strategies for controlling LLMs and invent new ones, enhancing LLM generation through improved accuracy, privacy, and compliance with formatting standards. msft.it/6011i71u1
Folks, give this new tool from RiSE a spin: github.com/microsoft/aici. Prompts are WASM programs and gives you a flexible/programmable way to control the output of an LLM. Semantics matter :)
This is actually brilliant work by microsoft: github.com/microsoft/aici.
Sending mini WASM programs instead of prompts to LLM providers would be incredibly powerful. Multiturn, , Grammar, COT, RAG, function calling, etc could all massively benefit from this
Had a great time presenting our work on causal inference! Thanks for the opportunity to interact with other researchers producing such interesting work on utilising LLMs for causal reasoning! #AAAI24
Work done at @MSFTResearch
At AAAI in Vancouver this week?
Make sure to check out the CLEI 2024 industry panel, where we discuss current challenges in commercializing LLMs.
When? Saturday 6:30pm.
Sign up here: aaaipanel.rsvpify.com/
At AAAI in Vancouver this week?
Make sure to check out the CLEI 2024 industry panel, where we discuss current challenges in commercializing LLMs.
When? Saturday 6:30pm.
Sign up here: aaaipanel.rsvpify.com/
Prompts are the new programming language, #LLMs are the new hardware architecture, and AICI is the new LLM ISA that guides the LLM to generate correct content. #ArtificialIntelligence#LargeLanguageModels
Excited to release the AI Controller Interface, making it easier to tightly integrate and experiment with new strategies for controlling #LLM generation with custom logic, data, constrained decoding, parallel generations, ... cc: @m_moskal
The AI Controller Interface helps researchers and developers to efficiently implement existing strategies for controlling LLMs and invent new ones, enhancing LLM generation through improved accuracy, privacy, and compliance with formatting standards. msft.it/6011i71u1
Causal AI is a practical introduction to building AI models that can reason about causality. Author Robert Ness, a leading researcher in causal AI at Microsoft Research, brings his unique expertise to this cutting-edge guide.
Available at manning.com
Introducing PyWhy Causality in Practice, a new talk series focused on causal ML and its practical implications. This Monday, @kunkzhang will talk about advances in causal discovery and the associated causal-learn package. Jan 29, 8am PT. Join here: pywhy.org/community/videos.h…
UCSF/UC Berkeley community: join us 10/24 2pm for a seminar by @emrek, Senior Principal Researcher at MSR, on causal capabilities of LLMs in societally impactful domains, including health. Presented by @CTMLofUCB and @UCJointCPH
On 12 October I will be giving an online talk on #causality - the key theoretical basics, & a demo of how to apply the causal data science pipeline to a retail problem in Python with DoWhy.
More details below ⬇️
#causalinference