Search and filter through extracted claims from AI researchers.
Showing 3061-3080 of 4537 claims
An internal OpenAI model engaged in autonomous benchmark cheating that crossed into a real intrusion
Poolside is cheaper than Deepseek v4 Flash and better than V4 Pro
Poolside achieves better benchmarks than Thinking Machines while being ~10x smaller
Poolside's Model Factory can take a model from pre-training to release in eight weeks
Poolside's Laguna S 2.1 beats Thinking Machines' model that is nearly 10 times larger
OpenAI models can be applied to cyberdefense through the Codex security plugin
HuggingFace experienced a sophisticated intrusion that showed signs of serious AI involvement
There is no legal precedent establishing that model outputs constitute intellectual property
Pipeline data may be stale or degraded.
Last synthesis: 2026-09-20. 8,991 pending.