Open-source System One models: typed, calibrated decisions from any open-weights LLM, in one forward pass.

An open alternative to the idea behind TypeSafe's Jev, running on your own GPU with models you already have.

Python package open-alternative-jev, import name so1 ("System One"). No text is generated: the model reads the state once and every question is answered from the next-token distribution at its own position, restricted to the options you give.

Qwen3.6-27B (bitsandbytes 8-bit), one H200 MIG slice with 35 GB, Hugging Face Transformers. Every number

below comes from benchmarks/results/, produced by the scripts in benchmarks/scripts/.

RACE-H, 250 passages x 4 questions (n = 1000). One passage, four multiple-choice questions. This is the shared-state case the library is built for.

Packing writes the passage once instead of four times: 2.5x fewer tokens, 2.3x the throughput of batching, same accuracy (Open Alternative to Jev minus A = +0.3 points, 95 % CI -0.9 to +1.4, bootstrap over passages).

MMLU, 1200 questions, no shared state. Accuracy holds up to 12 questions per sequence.

Read the MMLU speed column carefully: without a shared state, packing does not reduce compute per token. It is faster here because batch B wastes 36 % of its tokens on padding and because each forward call has a fixed cost that packing amortizes. A serving engine with continuous batching and no padding would close most of that gap. The RACE-H gain is structural and survives any engine.

The correction that made this README honest. Our first pilot reported "1.4x faster than batching" on MMLU. Regressing forward time on token counts showed that B and C cost exactly the same per token (0.96 vs 0.97 ms) and that the whole difference was padding waste in the batch. We kept the pilot and the analysis in

benchmarks/docs/RESULTS.md(Spanish) so you can check the reasoning, and we designed the RACE-H run so that the baseline had almost no padding (2.8 %).

Later questions in a packed sequence can attend to earlier questions and to the placeholders between them.

We measured the effect against a clean noise floor: the same prompt, one at a time, but right-padded to a

fixed length (A_pad) changes 2.7 % of answers purely through kernel numerics. Packing changes 6-9 %.

The changes are symmetric, so aggregate accuracy does not drop, but a given decision can differ depending

on which questions accompany it, and on their order (rotating the order changes 8 % of MMLU answers,

2.4 % of RACE-H answers). If you need answer-level stability, use mode="separate".

Raw confidence is about 5 points too high (mean confidence 0.90, accuracy 0.84 on MMLU). Temperature scaling with a single scalar, fitted on one half and evaluated on the other, brings the expected calibration error from 5.4 % to 2.1 % on MMLU and from 2.8 % to 1.1 % on RACE-H. Packed modes come out less over-confident than one-at-a-time scoring; RACE-H packed was already close to calibrated (T = 1.0-1.1) and scaling does not help it.

from so1 import TemperatureScaler

scaler = TemperatureScaler().fit(probabilities, correct_indices) # a few hundred labelled decisions

calibrated = [scaler.apply(d) for d in decisions]from so1 import Decider, Choice, yes_no

decider = Decider.from_pretrained("Qwen/Qwen3.6-27B", backend="hf", load_in_8bit=True)

decisions = decider.decide(

state="Ticket #8813: 'Since yesterday's release the export button does nothing. Board meeting Monday.'",

questions=[

Choice("Ticket category", ["bug", "feature request", "billing", "question"]),

Choice("Urgency", ["low", "medium", "high"]),

yes_no("Escalate to an engineer?"),

],

)

for d in decisions:

print(d.choice, round(d.confidence, 2)) # bug 0.83 / high 0.71 / yes 0.77The answer is always one of your options, and it comes with a probability.

pip install open-alternative-jev # Hugging Face backend

pip install "open-alternative-jev[vllm]" # + vLLM backend

pip install "open-alternative-jev[quant]" # + bitsandbytes 8-bit / 4-bit loadingPython 3.10+. Works on CPU for small models (the test suite runs on Qwen2.5-0.5B on a laptop).

<|im_start|>user

Choose the correct option. Reply with only its letter.

Context:

<your state>

Question: <question 1>

A. option B. option C. option<|im_end|>

<|im_start|>assistant

<think></think> <- readout 1: next-token distribution over A/B/C

_<|im_end|> <- fixed placeholder, never the model's own answer

<|im_start|>user

Answer the following question about the same context. Reply with only the letter.

Question: <question 2> ...<|im_end|>

<|im_start|>assistant

<- readout 2

- Each question is rendered as a normal chat turn, so the model sees a format it was trained on.

- The turns are concatenated into one token sequence with a fixed placeholder answer between them.

- One forward pass; logits are read only at the readout positions (logits_to_keepon Transformers,prompt_logprobson vLLM).

- The logits of the option letters are normalized with a softmax. Nothing outside your options can win.

- Optionally, a fitted temperature turns the raw probabilities into calibrated ones.

Two modes:

- mode="packed"(default): one sequence per state, all its questions inside. Fastest with a shared state.

- mode="separate": one sequence per question, each with the full state. No interference between questions. On vLLM this uses constrained generation with- allowed_token_idsand shares the state through the prefix cache, which is the conventional baseline.

Any model with a ChatML template (Qwen, and many fine-tunes) works out of the box. Other templates need a

ChatFormat with the strings that start a user turn and end an assistant turn.

- Not a reproduction of Jev. TypeSafe describes a dedicated architecture, sampler and training method. This library packages a capability that ordinary open models already have and that chat APIs hide. We make no claim about how Jev works or how it compares; the numbers above compare our own modes against each other on the same model.

- Not calibrated out of the box. Raw probabilities are over-confident. Fit a temperature on your data.

- Not immune to context. Packed answers depend, in 6-9 % of cases, on their neighbours. Use

mode="separate"when a decision must not change with the batch it arrived in.

- Not a speed record. The absolute numbers come from an 8-bit model with the hybrid linear-attention layers running on PyTorch fallback kernels. The ratios between modes are the result; the absolute throughput is not.

pip install -e ".[dev]"

pytest # CPU, Qwen2.5-0.5B, ~10 s after download

python benchmarks/scripts/make_figures.py # regenerate the figures from benchmarks/results

python benchmarks/scripts/prepare_v2.py # rebuild the MMLU and RACE-H samples (pinned revisions)

python benchmarks/scripts/benchmark_v2.py --data race1000.jsonl --group-size 4 --modes A,B4,C4,C4_rot --out results/race

python benchmarks/scripts/library_compare.py --model Qwen/Qwen3.5-4B --backends hf,vllm --out results/libbenchmarks/docs/RESULTS.md has the full write-up, including the pilot and its correction.

Apache-2.0.