Meta · LLM Test Generation
Catching JiTTests
Meta has moved Catching JiTTests into production, an LLM system that writes tests designed to fail on risky code diffs before they land. Per the page, it caught 74 privacy SEVs in H1 2026, with a 60% engineer-reported true-positive rate. The page does not state public availability, so it appears to be an internal tool.
Open the source ↗Similar Developer Tools releases
- NVIDIA Vera Rubin NVL72 in MLPerf Training v6.1
NVIDIA · Training Benchmark Results · Week 48, 2026
- FrontierFinance Benchmark Grader
Samaya AI · Finance agent benchmark · Week 47, 2026
- AWS Emulator
AWS · Agent-built AWS emulator · Week 47, 2026
- UAEF (Universal Agent Evaluation Framework)
AWS Machine Learning · Agent Eval Framework · Week 47, 2026
- Token Pooling-Aware Training for Multi-Vector Embedding Models
Hugging Face (Tom Aarsen) · Pooling-aware embeddings · Week 47, 2026
- Gemini Decisions API (Gemini Decision 1)
Google DeepMind · Multimodal Decision API · Week 47, 2026
More from Meta
- Muse Realtime Voice and Muse Realtime Avatar
Meta · Realtime voice avatars · Week 46, 2026
- Muse Spark Open
Meta · Open Weight Frontier · Week 43, 2026
- Muse Gadgets
Meta · Agent Hardware SDK · Week 41, 2026
- Muse
Meta · Personal AI Agent · Week 37, 2026
- Muse Glimmer
Meta · Reasoning · Week 33, 2026
- Muse Spark 1.2 & Muse Code
Meta · Coding Agents · Week 32, 2026
Published by Quanta AI Labs, an AI agency in India. Browse every week in AI News Weekly.