whatmodel.sh / decision / Zev (decision readout)

Zev (decision readout)

+ contribute

Zev (decision readout) is a source-available decision implementation from bhubbard.

Catalogue identity notes

Exact Decision Index tested engine zev-latest; reported model Zev. Type: inference technique; technique: autoregressive. Implementation: https://github.com/bhubbard/zev-rs. Comparison entry; no current run command, API price or speed inferred from the benchmark.

15.1±1.1#353 of 355 decision modelsfrom 42 benchmark results across 1 source

$ how it compares

The decision models ranked either side of Zev (decision readout), out of 357 with a score.

#	model	score	where	released
353	Nirnaya 0.4B kshitijthakkar	16±1	model files	2026-10-06
354	LFM2.5-350M-RLCD notnotsamuel	16±1	model files
355	Zev (decision readout) bhubbard	15±1	source code
356	gte-reranker-modernbert-base (DecisionBench f7481e60) Alibaba	7±2	model files
357	certo-decision-model (DecisionBench 7628c0da) altslate	7±2	model files

see all decision models · find the ones that fit your machine

$ what it's good at

tools & automation▁▃▁▁▁▂▃▄▆██▆ #249 of 250 · 10% of its score · 6 tests

Beats jeff GLiFormer-large (decision readout) here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: BFCL (case exact accuracy) submitted	32.5% ✎	▁▃▁▁▁▂▂▄▅▆█▁ beats 1/246	very easy · avg 92%	▇▇▇ 7.1%	sharp · at its level · noisy
Decision Index 0.4.1: API-Bank (accuracy) submitted	1.0% ✎	▁▃▁▁▁▁▁▂▃▂█▂ beats 1/236	easy · avg 70%	▇ 1.4%	sharp · noisy
Decision Index 0.4.1: When2Call (accuracy) submitted	14.7% ✎	▃▁▁▂▃▅▅▆█▇█▃ beats 1/246	easy · avg 72%	▇ 1.0%	at its level · noisy
Decision Index 0.4.1: ToolRet (nDCG@10) submitted	18.3% ✎	▁▁▁▃▁·▁▁▁▄█▁ beats 10/246	medium · avg 60%	▇ 0.8%	blunt · at its level
Decision Index 0.4.1: Home appliance simulator (case exact accuracy) submitted	0.0% ✎	█▁▁▃▄▄▆▄▁▂▁▃ beats 4/243	medium · avg 37%	▇ 0.0%	sharp · too hard for it · noisy
Decision Index 0.4.1: RouterBench (selected quality (quality objective)) submitted	68.6% ✎	no other model yet		—	failed run: Publisher excludes this protocol because its prompt reveals GPT-4 as the best route. Evidence: https://huggingface.co/spaces/multimodalart/jev-decision-index/blob/c130faf65f82e81d3bdefb01e758c712dc5c5199/data/index.json

retrieval & classification▁▁▃·▁▁▁▁▃▃█▅ #252 of 258 · 4.8% of its score · 7 tests

Beats DM-JEPA 1.1, JevK5 Lite, Qwen-2.5-1B-RLCD (decision readout) here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: Amazon ESCI (macro-F1) submitted	13.4% ✎	▁▁▁▃▁▁▂▂▃▃█▅ beats 6/246	medium · avg 47%	▇ 1.3%	typical
Decision Index 0.4.1: BRIGHT (nDCG@10) submitted	16.1% ✎	▁▁▁▃▁▁▁▂▂▄▇█ beats 10/246	medium · avg 38%	▇ 1.2%	blunt
Decision Index 0.4.1: HoVer (accuracy) submitted	50.0% ✎	▆▆▇▅▄▄▃▇▄█▆▄ beats 1/246	easy · avg 74%	▇ 0.7%	blunt · at its level
Decision Index 0.4.1: CLINC150+OOS (macro-F1) submitted	0.0% ✎	▃▁▁▁▁▁▁▂▃▆█▄ beats 0/238	easy · avg 83%	▇ 0.7%	sharp · at its level · noisy
Decision Index 0.4.1: PhishNChips phishing decisions (accuracy) submitted	49.6% ✎	█▅▅▄▃▂▁▁▁▁·▁ beats 8/245	easy · avg 70%	▇ 0.6%	at its level · noisy
Decision Index 0.4.1: BANKING77 (macro-F1) submitted	0.0% ✎	▃·▁▁▁▁▁▁▂█▇█ beats 0/238	easy · avg 79%	▇ 0.5%	at its level · noisy
Decision Index 0.4.1: SGD/SGD-X (macro-F1) submitted	40.0% ✎	no other model yet		—	failed run: Publisher reports a builder bug exposing dialogue turns after the classified turn; awaiting corrected rerun. Evidence: https://huggingface.co/spaces/multimodalart/jev-decision-index/blob/c130faf65f82e81d3bdefb01e758c712dc5c5199/data/index.json

language understanding▃▂▂▃▂▅▆▃█▆▄▃ #255 of 256 · 23% of its score · 9 tests

Beats Prumo (rule-based engine) here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: HellaSwag (accuracy) submitted	25.0% ✎	▃▄▂▂▄▇▃▇█▅▃▁ beats 2/246	very easy · avg 85%	▇▇▇ 8.4%	sharp
Decision Index 0.4.1: NLI4CT (macro-F1) submitted	40.0% ✎	▁▁▃▂▂▃▃▄▇▇█▄ beats 4/246	easy · avg 74%	▇▇ 6.2%	at its level
Decision Index 0.4.1: ANLI (macro-F1) submitted	16.7% ✎	▃▂▃▅▅█▂····▁ beats 1/246	medium · avg 60%	▇ 3.5%	typical
Decision Index 0.4.1: FinEntity (macro-F1) submitted	13.9% ✎	▁···▃▁▁▁▃▆█▃ beats 2/246	easy · avg 84%	▇ 1.2%	at its level · noisy
Decision Index 0.4.1: ContractNLI (macro-F1) submitted	10.8% ✎	▃▁▁▁▁▁▁▂▂▇▇█ beats 1/246	easy · avg 71%	▇ 1.0%	at its level · noisy
Decision Index 0.4.1: iSarcasmEval (Sarcasm F1 · track A, English) submitted	23.5% ✎	▁·▁·▁▁▂▄▅█▇▄ beats 30/246	medium · avg 44%	▇ 0.9%	typical
Decision Index 0.4.1: VAST (macro-F1) submitted	16.9% ✎	▃▁▁▂▅▄▄▂▄▆█▄ beats 0/246	medium · avg 62%	▇ 0.8%	at its level
Decision Index 0.4.1: RAGTruth (F1 on hallucinated class) submitted	41.5% ✎	▁▁▁▁▁▁▁▁▃▃▆█ beats 40/246	easy · avg 66%	▇ 0.6%	noisy

1 more test

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: ACOS (per-review F1) submitted	2.6% ✎	▁··▃▃▃▅▇█▇▃▃ beats 4/246	hard · avg 22%	▇ 0.2%	too hard for it · noisy

arts & human taste▁▇▄▃▅▅█▅▄▂▁▁ #234 of 247 · 7.5% of its score · 6 tests

Beats JEV Ettin 150M, JEV FLAN-T5 Large public v2 (342M), Kev 0.8B r15 here, though they rank higher overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: New Yorker (accuracy) submitted	21.0% ✎	▃▁▂▃▃▃▇▅█▅▃▁ beats 5/246	medium · avg 61%	▇ 2.7%	typical
Decision Index 0.4.1: BPoMP (accuracy) submitted	49.7% ✎	▁█▄▄▅█▆▅▄▁▁▁ beats 13/246	easy · avg 83%	▇ 2.0%	sharp · at its level · noisy
Decision Index 0.4.1: Humicroedit (accuracy) submitted	49.0% ✎	▁·▁▃▂▂▅█▄▂▁▁ beats 12/246	medium · avg 60%	▇ 1.6%	blunt · at its level
Decision Index 0.4.1: cfcolor (accuracy) submitted	50.2% ✎	▁········▁▇█ beats 18/246	medium · avg 60%	▇ 0.9%	blunt · at its level
Decision Index 0.4.1: Habermas Machine (accuracy) submitted	53.3% ✎	▁▂▄█▇▅▃▁▄▄▁▁ beats 182/246	medium · avg 47%	▇ 0.2%	blunt · at its level
Decision Index 0.4.1: POP909-CL (accuracy) submitted	1.5% ✎	▄▅▅▃▆▅▃▅▂▅▃█ beats 26/237	hard · avg 21%	▇ 0.0%	sharp · too hard for it · noisy

knowledge & reasoning▃▂▂▂▂▃▃▅▅▆█▂ #263 of 263 · 47% of its score · 14 tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: ARC-Easy (accuracy) submitted	26.0% ✎	▃▁▂▁▂▂▂▂▃▆█▅ beats 2/246	very easy · avg 96%	▇▇▇▇▇▇ 15%	sharp · at its level · noisy
Decision Index 0.4.1: ARC-Challenge (accuracy) submitted	23.2% ✎	▃▂▂▂▂▂▁▃▄▆█▆ beats 1/246	very easy · avg 90%	▇▇▇▇ 8.9%	sharp
Decision Index 0.4.1: MMLU (accuracy) submitted	22.9% ✎	▃▄▄▃▃▂▅▇▅█▇▁ beats 0/246	easy · avg 70%	▇▇ 5.5%	sharp
Decision Index 0.4.1: BBH (accuracy) submitted	32.6% ✎	▄▆▄▄█▇▇▄▃▁▁▁ beats 7/246	easy · avg 66%	▇▇ 4.6%	typical
Decision Index 0.4.1: WinoGrande (accuracy) submitted	49.6% ✎	█▅█▅▄█▆▄▃▃▁▁ beats 1/246	easy · avg 78%	▇▇ 4.1%	at its level
Decision Index 0.4.1: MMLU-Pro (accuracy) submitted	13.3% ✎	▄▄▃▃▂▄▅▇██▅▁ beats 10/246	medium · avg 46%	▇ 2.9%	sharp
Decision Index 0.4.1: GPQA Diamond (accuracy) submitted	24.5% ✎	▂▆▅▅▆█▆▃▁··▁ beats 12/246	medium · avg 39%	▇ 1.7%	blunt
Decision Index 0.4.1: MuSR (accuracy) submitted	35.8% ✎	▃▁▂█▅▄▁▁▁▁▁▁ beats 2/246	medium · avg 60%	▇ 1.0%	blunt · at its level

6 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: CLadder (accuracy) submitted	49.9% ✎	▁███▆▅▃▃▁▁▁▁ beats 8/246	easy · avg 71%	▇ 1.0%	at its level
Decision Index 0.4.1: GSM8K (accuracy (4- and 10-choice)) submitted	17.8% ✎	▁··▄█▆▆▆▅▅▃▁ beats 11/246	medium · avg 45%	▇ 0.8%	sharp · too hard for it
Decision Index 0.4.1: CRUXEval (accuracy) submitted	38.1% ✎	██▄▃▃▄▁▁···▁ beats 41/246	medium · avg 58%	▇ 0.7%	noisy
Decision Index 0.4.1: SATA-Bench (case exact accuracy) submitted	0.0% ✎	▃▁▁▁▁▁▁▂▅█▂▁ beats 4/246	hard · avg 22%	▇ 0.3%	too hard for it
Decision Index 0.4.1: ChessBench (accuracy) submitted	7.9% ✎	▄▇▄▄▄▆▄█▄▁▁▁ beats 19/246	hard · avg 16%	▇ 0.2%	blunt · too hard for it
Decision Index 0.4.1: HLE (accuracy) submitted	14.8% ✎	▁▁▂▃▆█▇▅▅▃▁▁ beats 189/246	very hard · avg 13%	▇ 0.0%	blunt

other tests2 not in a task · 6.8% of its score

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: private same skills (equated points)	0.7600 points ✎	▃▂▂▃▃▃▄▅▅▇█▃ beats 0/247	avg 41.206 points	▇▇ 4.4%	blunt
Decision Index 0.4.1: private new domains (equated points)	7.2900 points ✎	▃█▃▃▂▆▃▃▄▄▄▄ beats 7/247	avg 38.861 points	▇ 2.4%	blunt

$ where to run it

Implementation and setup instructions ↗
Published implementation and setup documentation verified. This entry describes a decision readout or evaluated configuration, not necessarily separate trained weights. Follow the referenced version and model settings; hardware compatibility has not been verified. Source · checked 2026-10-31.

Model repository and documentation: github.com/bhubbard ↗

$ how the score was worked out · Decision Index 0.4.1 99%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

source	results	share of the score	licence
Decision Index 0.4.1 · submitted; independent execution unverified	42	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 99%	CC BY 4.0
prior · pull toward the middle		▇ 1%
weighted average		= 15.1

not counted (2)

Decision Index 0.4.1	Decision Index 0.4.1: RouterBench (selected quality (quality objective)) [correct]	68.6%	failed run: Publisher excludes this protocol because its prompt reveals GPT-4 as the best route. Evidence: https://huggingface.co/spaces/multimodalart/jev-decision-index/blob/c130faf65f82e81d3bdefb01e758c712dc5c5199/data/index.json
Decision Index 0.4.1	Decision Index 0.4.1: SGD/SGD-X (macro-F1) [correct]	40.0%	failed run: Publisher reports a builder bug exposing dialogue turns after the classified turn; awaiting corrected rerun. Evidence: https://huggingface.co/spaces/multimodalart/jev-decision-index/blob/c130faf65f82e81d3bdefb01e758c712dc5c5199/data/index.json

$ improve this page

Update model information Add a place to run it Add a test result Propose a benchmark Correct a benchmark score Edit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated 2026-11-03 02:37 UTC · sources and licences · json