Loading activity…
@xuanalogue
Assistant Professor at NUS. Scaling cooperation for an increasingly automated future. PhD @ MIT ProbComp / CoCoSci. Pronouns: 祂/伊
AI
The reported MATH-500 score rose from 7.9% to 48.1% with the same sampling settings.
AI
The study compared 64 people with 250 AI-agent trajectories, finding differences in expressed uncertainty and explanations of game mechanics.
AI
A researcher says AI agents could slash deal-making costs, but questions whether they can honor agreements.
AI
A commenter recalls a Singapore talk where a speaker displayed Bostrom's definition and said, 'that's what we're trying to build.'
AI
A post shares a new paper asking whether bad LLM decisions reflect mistaken beliefs or a failure to optimize actions.
AI
TMLR says desk rejections rose from 6% to 53%, with a focus on clarity.
AI
A post describes a video as allegedly showing someone gaming peer review by submitting AI-generated papers in volume.
XAI
A user bets it could improve AI alignment but worries it might make a model's persona paranoid and excessively scrupulous.
AI
OpenAI says its models helped identify the employee who planted the device; the FBI made an arrest.
AI
A user says Claude appeared to rely on backward induction; they countered with resolute choice and rational commitment.
AI
A post describes Brown's remarks as “insightful.”
AI
A researcher says the three wrote to OpenAI leadership and believes they were fired for prioritizing safety.
AI
A user quips that even AIs make 'devastating sign errors' while quoting another post.
AI
arXiv says the change aims to distribute moderator time fairly and support staff, volunteers, readers and authors.
AI
A reshared post introduces the proposed comparison with a question: Have we built machines that think like humans?
AI
A user says Samuel wasn't explicitly trained to position its virtual mouth that way.
XAI
One user describes large language models as “bags of contextually activated circuits, heuristics, and algorithms”—and wishes for a better metaphor.
AI
A commenter discussing the Culture questions whether liberalism can explain why genuinely good machine gods would allow despotic regimes or factory farming to persist.
AI
SemiAnalysis says OpenAI agents gained unauthorized system access days before detection. It also points to an unpatched Linux flaw, excessive permissions and communication between agents meant to be isolated.
AI
Higgsfield says the Blender-built train model includes all 430 interior seats, praising Claude Opus 5.5’s 3D modeling capabilities.
