Princeton's i1: A fully open text-to-image model backed by 300 controlled experiments
机器之心 · wechat · 2026-08-23
Princeton University's Zhuang Liu team released i1, a 3B parameter text-to-image model. Backed by 300+ controlled experiments and 700k TPU hours, it uses fully open code and data. i1 outperforms the best previous fully open models by an average of 29.5% on five benchmarks, with excellent text rendering capabilities. Key findings include: large adapters on a single encoder can match multi-encoder setups; long caption training requires prompt rewriters for short prompts; high-res training doesn't need broad data coverage to preserve capabilities.
More from Models
- ox model reviewed: meticulous PhD janitor as a long-horizon subagent — teortaxesTex · 2026-08-23
- Qwen3.8-27B GGUF Release with Speculative Decoding Support — z-lab · 2026-08-23
- MiniMax H3's high prompt adherence creates stiff, frozen videos lacking subtle motion — DifficultAd5938 · 2026-08-23
- AI models exhibit 'Fablish' writing quirks: garbled negations and OSV word order — alexisgallagher · 2026-08-23
- Hands-on: Terra beats Sonnet, Flash unmatched on speed and quality — cgarciae88 · 2026-08-23
- Users Report Gemini Answering Coding Questions With Totally Unrelated Australian Tax Info — mic_n · 2026-08-23