Loading activity…
@willdepue
dei ex machina prev @ openai (sora 1 & 2, posttraining o3/4o, pretraining moonshots)
AI
The reported MATH-500 score rose from 7.9% to 48.1% with the same sampling settings.
AI
Astra's claimed robotics edge led one observer to speculate about transfer at scale; a reply on X questioned the surprise.
AI
AlexKsendzovsky claims a starting price of 0.5¢ per second and a speed advantage over an unnamed base model.
XAI
The user says they're on Muse Wonder's Power plan and tagged Muse to flag the incident.
AI
The introductory post says users can get highlights from footage, edit videos and share links for feedback.
XAI
willdepue's follow-up quote-post linked another video but offered only the poster's account of the Codex result.
AI
Unlike models with fixed tokenizers, H-Net learns to group raw bytes within the model, Cartesia says.
XAI
The Washington Post reports the researchers used AI agents to access Slack channels discussing Anthropic’s most sensitive models.
AI
One user says Claude returns great results when left for 12 hours, while Codex's app keeps breaking.
AI
A user says the interruptions do not restart their overnight sessions and asks which models remain available.
AI
One user says waiting for generated code without knowing what the agent is doing is exhausting.
AI
The contest page lays out GPU-kernel and SGLang serving work, with different tests for each.
AI
Tesla's report covers about 3.1 million autonomous miles through Oct. 18, 2026, excluding miles with an onboard safety monitor.
AI
In a post, kushal1t claims the framework boosts GPU utilization by packing independent workloads into one kernel.
AI
Anthropic reports 17 of 211 lab-tested designs met binding criteria, but none were tested for toxin neutralization.
AI
A commenter says leaving for a competitor can bring a raise and an invitation to return.
AI
0x00B1's October 19 post tagged @ArfurGrok but did not mention a role at OpenAI.
AI
A user asks how models could learn to do this and points to a thread.
AI
A post says looping at a given noise level also reduces reliance on CFG.
AI
A user compares AI agents to voices layered over consciousness, then jokes about meditating both away.
