Qwen3.5-4B with a merged rank-16 LoRA distilled from 17,408 teacher questions and 30,000 public training rows, read out as a temperature-scaled softmax over answer-letter logits. Also in 9B, 2B and GGUF.
A 144M-parameter System One model on the multilingual mmBERT-small encoder. Takes a state, a question and 2 to 20 options and returns one decision with probabilities for choice, score and yes/no questions, on a CPU.
Decides
choice, score, noul, classify, route
choice, score, noul, classify, route
Architecture
jevk5
julia
Fine-tuned from
qwen/qwen3.5-4b
jhu-clsp/mmbert-small
License
apache-2.0
apache-2.0
Availability
Open weights
Open weights
Hosted by
—
—
Input price
—
—
Decision accuracy
78.4%
73.2%
Calibration error
0.035
—
Valid action rate
—
—
Median latency
13.2 ms
—
p95 latency
—
—
Evaluation suite
JevBench public hard tier
Figures are from each model’s manifest; accuracy and latency are what the publishers report, on their own suites and hardware. Add a third model.
Questions
What is the difference between jevk5 and julia-1?
jevk5 is from Alibi Serikbay and julia-1 from Supersonic Labs. Both have open weights you can download and run. Both answer choice, score, noul, classify and route questions. julia-1 is the smaller model, at 144M parameters to 4.0B.
Which is more accurate, jevk5 or julia-1?
They report on different suites — jevk5 78.4% on JevBench public hard tier, julia-1 73.2% on typed-decisions test set (400 cases, 2,000 questions) — so the numbers do not rank them. Test both on your own labelled examples.
Which is cheaper, jevk5 or julia-1?
jevk5: Free (open weights). julia-1: Free (open weights). Open weights cost nothing per call beyond your own hardware.
Can I run jevk5 or julia-1 locally?
Yes, both: systemone pull alibi-serikbay/jevk5 and systemone pull supersonic-labs/julia-1 download the weights.
typed-decisions test set (400 cases, 2,000 questions)
