Shane Larson PRO
comgen42
AI & ML interests
None yet
Recent Activity
updated a model 32 minutes ago
cortex-agent-llc/kodiak-v0.3-1b-accuracy posted an update 43 minutes ago
š» Kodiak-v0.3-1B: seven new kinds of decision, measured on real data.
Kodiak is an open 1B encoder that answers typed questions with calibrated confidence, or says "can't tell". New in v0.3:
ā¢ pairwise judge (which answer is better?)
ā¢ long-answer hallucination checks
ā¢ stance and sarcasm
ā¢ policy violation and refund-eligibility checks against your written rules
ā¢ agent step safety (run it, ask first, or never)
We stopped trusting our own synthetic tests and judged v0.3 on real labelled data it never trained on (RAGBench, MT-Bench human judgments, SemEval stance): 0.21 ā 0.36, averaged over 3 training runs. Nothing else got worse, and when it says "can't tell" it's right 93% of the time (up from 88%).
Known limits are in the model card.
š¦ https://huggingface.co/cortex-agent-llc/kodiak-v0.3-1b
šÆ Accuracy mode: https://huggingface.co/cortex-agent-llc/kodiak-v0.3-1b-accuracy
š¹ļø Demo: https://huggingface.co/spaces/comgen42/kodiak-demo
š What we built, and the kill that changed our process: https://cortexagent.com/blog/kodiak-v0-3-seven-new-kinds-of-decision-measured-on-real-data
updated a Space about 1 hour ago
comgen42/kodiak-demoOrganizations
