Inhwa Song
@inhwa.bsky.social
Incoming PhD student @Princeton CS, Currently ugrad @KAIST CS, Prev NAVER AI Lab, Georgia Tech
inhwasong.com
inhwasong.com
Reposted by Inhwa Song
Do reasoning models have real “Aha!” moments—mid-chain realizations where they intrinsically self-correct?
In a new pre-print, “The Illusion of Insight in Reasoning Models," led by @liv-daliberti.bsky.social we provide strong evidence that they do not!
📜: arxiv.org/abs/2601.00514
In a new pre-print, “The Illusion of Insight in Reasoning Models," led by @liv-daliberti.bsky.social we provide strong evidence that they do not!
📜: arxiv.org/abs/2601.00514
January 5, 2026 at 7:39 PM
Do reasoning models have real “Aha!” moments—mid-chain realizations where they intrinsically self-correct?
In a new pre-print, “The Illusion of Insight in Reasoning Models," led by @liv-daliberti.bsky.social we provide strong evidence that they do not!
📜: arxiv.org/abs/2601.00514
In a new pre-print, “The Illusion of Insight in Reasoning Models," led by @liv-daliberti.bsky.social we provide strong evidence that they do not!
📜: arxiv.org/abs/2601.00514
Reposted by Inhwa Song
help spread the word about this campaign - or, if you're able, consider supporting yourself! Bonfire is building a modular toolkit for organizing, mutual aid, open science, and collective decision-making on the open social web. communities choose what gets built:
www.indiegogo.com/en/projects/...
www.indiegogo.com/en/projects/...
Bonfire: building blocks for communities on the open social web by Bonfire - Indiegogo
From platforms to communities: a federated social web with shared foundations and co‑created spaces for gathering, care, connection, and stewardship.
www.indiegogo.com
December 3, 2025 at 6:15 PM
help spread the word about this campaign - or, if you're able, consider supporting yourself! Bonfire is building a modular toolkit for organizing, mutual aid, open science, and collective decision-making on the open social web. communities choose what gets built:
www.indiegogo.com/en/projects/...
www.indiegogo.com/en/projects/...
Independent of the societal debates of using LLM chatbots for mental health support, people experiencing distress continue to use them.
I’ll be talking about this at #CSCW2025 — AI Applications for Safety and Support 🗓️ Tue, Oct 21 | 14:30 PM
Paper: dl.acm.org/doi/10.1145/...
I’ll be talking about this at #CSCW2025 — AI Applications for Safety and Support 🗓️ Tue, Oct 21 | 14:30 PM
Paper: dl.acm.org/doi/10.1145/...
The Typing Cure: Experiences with Large Language Model Chatbots for Mental Health Support | Proceedings of the ACM on Human-Computer Interaction
People experiencing severe distress increasingly use Large Language Model (LLM) chatbots as mental health support tools. Discussions on social media have described how engagements were lifesaving for some, but evidence suggests that general-purpose LLM ...
dl.acm.org
October 21, 2025 at 8:46 AM
Independent of the societal debates of using LLM chatbots for mental health support, people experiencing distress continue to use them.
I’ll be talking about this at #CSCW2025 — AI Applications for Safety and Support 🗓️ Tue, Oct 21 | 14:30 PM
Paper: dl.acm.org/doi/10.1145/...
I’ll be talking about this at #CSCW2025 — AI Applications for Safety and Support 🗓️ Tue, Oct 21 | 14:30 PM
Paper: dl.acm.org/doi/10.1145/...
Reposted by Inhwa Song
Social media feeds today are optimized for engagement, often leading to misalignment between users' intentions and technology use.
In a new paper, we introduce Bonsai, a tool to create feeds based on stated preferences, rather than predicted engagement.
arxiv.org/abs/2509.10776
In a new paper, we introduce Bonsai, a tool to create feeds based on stated preferences, rather than predicted engagement.
arxiv.org/abs/2509.10776
September 16, 2025 at 1:24 PM
Social media feeds today are optimized for engagement, often leading to misalignment between users' intentions and technology use.
In a new paper, we introduce Bonsai, a tool to create feeds based on stated preferences, rather than predicted engagement.
arxiv.org/abs/2509.10776
In a new paper, we introduce Bonsai, a tool to create feeds based on stated preferences, rather than predicted engagement.
arxiv.org/abs/2509.10776
Reposted by Inhwa Song
Media | KAIST STP
STP, Graduate School of Science and Technology Policy
stp.kaist.ac.kr
May 19, 2025 at 5:23 AM