Open-source Laya classifies with calibrated probabilities in a single 33ms pass
minchoi · x · 2026-09-22
Laya is an open-source multilingual, non-autoregressive System 1 decision model: given a state (text, email, ticket, or JSON) and typed questions, it returns typed answers with mathematically calibrated probabilities in one forward pass (33ms) across 100+ languages. It's trained with RLCD against strictly proper scoring rules, so honest probabilities maximize reward—and since it never generates text, there's nothing to parse and nothing to hallucinate. Three checkpoints ship: English (ModernBERT-large, 421M, guardrails/email triage), multilingual (mmBERT-base, 322M, 2.2x faster), and typed-decisions (421M, 0.766 accuracy), with a built-in Router that detects scripts/languages in sub-milliseconds and dispatches to the optimal checkpoint.
Related event: Open-Source Laya Delivers Zero-Hallucination Decisions in 33ms(2 posts)→
More from Models
- Claude raises five-hour limits and adds a save-for-later rate limit reset — EricBuess · 2026-09-23
- Claude Opus 5.5 launches: Fable 5.1-level performance at 40% less cost — EricBuess · 2026-09-23
- Claude Opus 5.5 Benchmark Scores Circulating on Reddit — FalconsArentReal · 2026-09-23
- Anthropic Launches Claude Opus 5.5: Smarter and 40% Cheaper — ResultBackground2450 · 2026-09-23
- Sparse distillation paper: supervising just 0.1%-1% of tokens can match or beat full OPD — jiank_uiuc · 2026-09-23
- Distillation's real impact on Chinese labs debated: no hard evidence, says Lambert, maybe 1-2 month edge — xeophon · 2026-09-23