Hongli Zhan ✈️ ICML
@hongli-zhan.bsky.social
http://honglizhan.github.io
PhD Candidate 🤘@UTAustin | previously @IBMResearch @sjtu1896 | NLP for social good
PhD Candidate 🤘@UTAustin | previously @IBMResearch @sjtu1896 | NLP for social good
Pinned
I'll be at #ICML to present SPRI next week! Come by our poster on Tuesday, July 15, 4:30pm, and let’s catch up on LLM alignment! 😃
🚀TL;DR: We introduce Situated-PRInciples (SPRI), a framework that automatically generates input-specific principles to align responses — with minimal human effort.
🧵
🚀TL;DR: We introduce Situated-PRInciples (SPRI), a framework that automatically generates input-specific principles to align responses — with minimal human effort.
🧵
I'll be at #ICML to present SPRI next week! Come by our poster on Tuesday, July 15, 4:30pm, and let’s catch up on LLM alignment! 😃
🚀TL;DR: We introduce Situated-PRInciples (SPRI), a framework that automatically generates input-specific principles to align responses — with minimal human effort.
🧵
🚀TL;DR: We introduce Situated-PRInciples (SPRI), a framework that automatically generates input-specific principles to align responses — with minimal human effort.
🧵
July 8, 2025 at 3:05 PM
I'll be at #ICML to present SPRI next week! Come by our poster on Tuesday, July 15, 4:30pm, and let’s catch up on LLM alignment! 😃
🚀TL;DR: We introduce Situated-PRInciples (SPRI), a framework that automatically generates input-specific principles to align responses — with minimal human effort.
🧵
🚀TL;DR: We introduce Situated-PRInciples (SPRI), a framework that automatically generates input-specific principles to align responses — with minimal human effort.
🧵
I’m excited to share that our paper has been accepted at #ICML2025! 🎉🥳🎊
This work was done during my internship at IBM Research, and it wouldn’t have been possible without a top-notch team and my amazing advisor 👏
This work was done during my internship at IBM Research, and it wouldn’t have been possible without a top-notch team and my amazing advisor 👏
May 2, 2025 at 9:27 PM
I’m excited to share that our paper has been accepted at #ICML2025! 🎉🥳🎊
This work was done during my internship at IBM Research, and it wouldn’t have been possible without a top-notch team and my amazing advisor 👏
This work was done during my internship at IBM Research, and it wouldn’t have been possible without a top-notch team and my amazing advisor 👏
Reposted by Hongli Zhan ✈️ ICML
To appear #ICML2025!! 🎉
Constitutional AI works great for aligning LLMs, but the principles can be too generic to apply.
Can we guide responses with context-situated principles instead?
Introducing SPRI, a system that produces principles tailored to each query, with minimal to no human effort.
arxiv.org/pdf/2502.03397
Can we guide responses with context-situated principles instead?
Introducing SPRI, a system that produces principles tailored to each query, with minimal to no human effort.
arxiv.org/pdf/2502.03397
May 2, 2025 at 12:34 PM
To appear #ICML2025!! 🎉
Reposted by Hongli Zhan ✈️ ICML
The principles that LLMs align with should be specific to the task at hand! Check out @hongli-zhan.bsky.social’s latest work 👇
Constitutional AI works great for aligning LLMs, but the principles can be too generic to apply.
Can we guide responses with context-situated principles instead?
Introducing SPRI, a system that produces principles tailored to each query, with minimal to no human effort.
arxiv.org/pdf/2502.03397
Can we guide responses with context-situated principles instead?
Introducing SPRI, a system that produces principles tailored to each query, with minimal to no human effort.
arxiv.org/pdf/2502.03397
February 6, 2025 at 10:52 PM
The principles that LLMs align with should be specific to the task at hand! Check out @hongli-zhan.bsky.social’s latest work 👇
Constitutional AI works great for aligning LLMs, but the principles can be too generic to apply.
Can we guide responses with context-situated principles instead?
Introducing SPRI, a system that produces principles tailored to each query, with minimal to no human effort.
arxiv.org/pdf/2502.03397
Can we guide responses with context-situated principles instead?
Introducing SPRI, a system that produces principles tailored to each query, with minimal to no human effort.
arxiv.org/pdf/2502.03397
February 6, 2025 at 10:43 PM
Constitutional AI works great for aligning LLMs, but the principles can be too generic to apply.
Can we guide responses with context-situated principles instead?
Introducing SPRI, a system that produces principles tailored to each query, with minimal to no human effort.
arxiv.org/pdf/2502.03397
Can we guide responses with context-situated principles instead?
Introducing SPRI, a system that produces principles tailored to each query, with minimal to no human effort.
arxiv.org/pdf/2502.03397