v1.1, October 2026. Read-or-Think: one-pass answer, a short thought only when unsure. 4B, 9B, 27B.
AI & ML interests
None defined yet.
Recent Activity
View all activity
Papers
StartLux-Decision: AutoThink Decision Models with Reinforcement Learning for Calibrated Deliberative Decisions
Massive Activations in Hybrid Linear Attention Large Language Models: Pre-Attention Spikes and Inter-Spike Plateaus
models 65
startlux-models/ReaDy-Qwen3.8-27B
2B • Updated • 32 • 4
startlux-models/StartLux-Decision-v1.1-9B-Instant
10B • Updated • 301 • 1
startlux-models/StartLux-Decision-v1.1-9B-AutoThink
9B • Updated • 293 • 3
startlux-models/StartLux-Decision-v1.1-4B-Instant
5B • Updated • 439 • 1
startlux-models/StartLux-Decision-v1.1-4B-AutoThink
5B • Updated • 413 • 2
startlux-models/StartLux-Decision-v1.1-2B-Instant
2B • Updated • 256 • 1
startlux-models/StartLux-Decision-v1.1-27B-Instant
28B • Updated • 359 • 7
startlux-models/StartLux-Decision-v1.1-27B-AutoThink
27B • Updated • 412 • 9
startlux-models/StartLux-Decision-v1.1-0.8B-Instant
0.9B • Updated • 317 • 1
startlux-models/StartLux-Decision-v1.1-9B-Instant-MLX-8bit
9B • Updated • 70