Kas 4B

+ contribute

Kas 4B is an open-weights decision model from kpiya. Its published size is 4.02B parameters. Licence: apache-2.0.

Catalogue identity notes

Exact Decision Index tested engine kas-4b; reported model Kas 4B. Type: LoRA; technique: autoregressive. Comparison entry; no current run command, API price or speed inferred from the benchmark. Current repository metadata checked separately; an unpinned alias is not evidence that today’s weights reproduce this historical run.

Model information sources

· Weights licence ↗ · Parameter count ↗

Publisher parameter count; includes the served backbone, not just adapter size.

46.1±1.2#235 of 355 decision modelsfrom 42 benchmark results across 1 source

$ how it compares

The decision models ranked either side of Kas 4B, out of 357 with a score.

#	model	score	where
235	LatentMesh Blitz 2B v1 LatentMesh Labs	46±1	request access
236	Winnow-E4B EldanRing	46±1	model files
237	Kas 4B kpiya	46±1	model files
238	OpenThai-SystemOne (JevBench mini) iApp Technology	46±2	source code
239	vera-core-2 scar-ai	46±1	model files

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

$ what it's good at

tools & automation▁▃▁▁▁▂▃▄▆██▆ #159 of 250 · 8.5% of its score · 6 tests

Beats Trio-Spark DI (9B), LiquidAI d1-3B, vLLM-SR Decision 2.0 Nox 4B here, though they rank higher overall. Trails JevK5 Lite, Gevva E2B, Amber 0.4B here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: BFCL (case exact accuracy)	92.2% ✎	▁▂▁▁▁▂▂▄▅▆█▁ beats 94/246	very easy · avg 92%	▇ 3.2%	sharp · noisy
Decision Index 0.4.1: API-Bank (accuracy)	59.3% ✎	▁▁▁▁▁▁▁▂▃▂█▂ beats 68/236	easy · avg 70%	▇ 2.5%	sharp · at its level · noisy
Decision Index 0.4.1: When2Call (accuracy)	45.4% ✎	▁▁▁▂▃▅▅▆█▇█▃ beats 33/246	easy · avg 72%	▇ 1.1%	at its level · noisy
Decision Index 0.4.1: ToolRet (nDCG@10)	55.8% ✎	▁▁▁▁▁·▁▁▃▄█▁ beats 44/246	medium · avg 60%	▇ 0.9%	blunt · at its level
Decision Index 0.4.1: Home appliance simulator (case exact accuracy)	10.2% ✎	█▁▁▃▄▄▆▄▁▂▁▃ beats 68/243	medium · avg 37%	▇ 0.8%	sharp · at its level · noisy
Decision Index 0.4.1: RouterBench (selected quality (quality objective))	69.2% ✎	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▁▁▁·▁▁▁▁▃▃█▅ #112 of 258 · 5.7% of its score · 7 tests

Beats AutoTrust-JEV-27B v1.1, Quyet-1.0-Large, OneJev1.6-9B here, though they rank higher overall. Trails vLLM-SR vela3-fast (Vela 3.0, 0.4B), crystal-0.8b here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: BRIGHT (nDCG@10)	34.0% ✎	▁▁▁▁▁▁▁▂▃▄▇█ beats 66/246	medium · avg 38%	▇ 1.9%	blunt · at its level
Decision Index 0.4.1: Amazon ESCI (macro-F1)	55.1% ✎	▁▁▁▁▁▁▂▂▃▃█▅ beats 153/246	medium · avg 47%	▇ 1.9%	at its level
Decision Index 0.4.1: HoVer (accuracy)	63.1% ✎	▆▆▇▅▄▄▃▇▄█▆▄ beats 70/246	easy · avg 74%	▇ 0.6%	blunt · at its level
Decision Index 0.4.1: PhishNChips phishing decisions (accuracy)	53.9% ✎	█▅▅▄▃▂▁▁▁▁·▁ beats 65/245	easy · avg 70%	▇ 0.5%	at its level · noisy
Decision Index 0.4.1: CLINC150+OOS (macro-F1)	73.4% ✎	▁▁▁▁▁▁▁▂▃▆█▄ beats 47/238	easy · avg 83%	▇ 0.5%	sharp · noisy
Decision Index 0.4.1: BANKING77 (macro-F1)	59.4% ✎	▁·▁▁▁▁▁▁▃█▇█ beats 20/238	easy · avg 79%	▇ 0.3%	noisy
Decision Index 0.4.1: SGD/SGD-X (macro-F1)	67.5% ✎	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▂▂▂▃▂▅▆▃█▆▄▃ #180 of 256 · 24% of its score · 9 tests

Beats Nari Decision 4B v3.2, Solar Decide Flash, Valen-4B here, though they rank higher overall. Trails DecisionTune 1.0, vera-spark, jdi-ptr14-0.8b here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: HellaSwag (accuracy)	77.8% ✎	▂▄▂▂▄▇▃▇█▅▃▁ beats 76/246	very easy · avg 85%	▇▇▇ 7.5%	sharp
Decision Index 0.4.1: NLI4CT (macro-F1)	71.7% ✎	▁▁▁▂▂▃▃▄▇▇█▄ beats 83/246	easy · avg 74%	▇▇ 5.4%	at its level
Decision Index 0.4.1: ANLI (macro-F1)	54.5% ✎	▁▂▃▅▅█▂····▁ beats 81/246	medium · avg 60%	▇▇ 4.9%	at its level
Decision Index 0.4.1: iSarcasmEval (Sarcasm F1 · track A, English)	30.8% ✎	▁·▁·▁▁▂▄▅█▇▄ beats 62/246	medium · avg 44%	▇ 1.9%	at its level
Decision Index 0.4.1: ContractNLI (macro-F1)	50.3% ✎	▁▁▁▁▁▁▃▂▂▇▇█ beats 33/246	easy · avg 71%	▇ 1.0%	at its level · noisy
Decision Index 0.4.1: VAST (macro-F1)	52.8% ✎	▁▁▁▂▅▄▄▂▄▆█▄ beats 87/246	medium · avg 62%	▇ 0.9%	at its level
Decision Index 0.4.1: FinEntity (macro-F1)	82.3% ✎	▁···▁▁▁▁▃▆█▃ beats 81/246	easy · avg 84%	▇ 0.8%	noisy
Decision Index 0.4.1: RAGTruth (F1 on hallucinated class)	57.6% ✎	▁▁▁▁▁▁▁▁▁▃▆█ beats 66/246	easy · avg 66%	▇ 0.8%	at its level · noisy

1 more test

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: ACOS (per-review F1)	10.5% ✎	▁··▂▃▃▅▇█▇▃▃ beats 61/246	hard · avg 22%	▇ 0.7%	noisy

arts & human taste▁▇▄▃▅▅█▅▄▂▁▁ #27 of 247 · 8.4% of its score · 6 tests

Beats Jeffan-31b-1710, vLLM-SR d3-turbo v1.1 (27B), WeMM-JEV-27B 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)	50.4% ✎	▁▁▂▃▃▃▇▅█▅▃▁ beats 58/246	medium · avg 61%	▇ 3.7%	at its level
Decision Index 0.4.1: Humicroedit (accuracy)	60.4% ✎	▁·▁▃▂▂▅█▄▂▁▁ beats 101/246	medium · avg 60%	▇ 1.6%	blunt · at its level
Decision Index 0.4.1: BPoMP (accuracy)	97.0% ✎	▁█▄▄▅█▆▅▄▁▁▁ beats 222/246	easy · avg 83%	▇ 1.5%	sharp · noisy
Decision Index 0.4.1: cfcolor (accuracy)	64.7% ✎	▁········▁▇█ beats 169/246	medium · avg 60%	▇ 0.9%	blunt · at its level
Decision Index 0.4.1: POP909-CL (accuracy)	0.4% ✎	▄▅▅▃▆▅▃▅▂▅▃█ beats 1/237	hard · avg 21%	▇ 0.5%	sharp · noisy
Decision Index 0.4.1: Habermas Machine (accuracy)	70.0% ✎	▁▂▄█▇▅▁▁▄▄▁▃ beats 244/246	medium · avg 47%	▇ 0.3%	blunt · at its level

knowledge & reasoning▁▂▂▂▂▃▃▅▅▆█▂ #170 of 263 · 46% of its score · 14 tests

Beats Solomon v1.1, Standard One 3B SH (v4), baselabs joey-1 (2B) here, though they rank higher overall. Trails openvons Qwen3-4B (decision readout), Jevstral 8B, Jobe Qwen3.5-4B (decision readout) here, though they rank lower overall.

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: MMLU-Pro (accuracy)	42.3% ✎	▄▄▃▃▂▄▅▇██▅▁ beats 87/246	medium · avg 46%	▇▇▇ 8.5%	sharp · at its level
Decision Index 0.4.1: MMLU (accuracy)	68.6% ✎	▁▄▄▃▃▂▅▇▅█▇▁ beats 92/246	easy · avg 70%	▇▇▇ 7.4%	sharp · at its level
Decision Index 0.4.1: ARC-Challenge (accuracy)	88.9% ✎	▂▂▂▂▂▂▁▃▄▆█▆ beats 90/246	very easy · avg 90%	▇▇▇ 6.6%	sharp
Decision Index 0.4.1: BBH (accuracy)	57.0% ✎	▄▆▄▄█▇▇▄▃▁▁▁ beats 75/246	easy · avg 66%	▇▇▇ 6.3%	at its level
Decision Index 0.4.1: ARC-Easy (accuracy)	95.9% ✎	▁▁▂▁▂▂▂▂▃▆█▅ beats 89/246	very easy · avg 96%	▇▇ 4.2%	sharp · maxed out · noisy
Decision Index 0.4.1: WinoGrande (accuracy)	67.3% ✎	█▅█▅▄█▆▄▃▃▁▁ beats 67/246	easy · avg 78%	▇ 3.4%	at its level
Decision Index 0.4.1: GSM8K (accuracy (4- and 10-choice))	25.8% ✎	▁··▄█▆▆▆▅▅▃▁ beats 61/246	medium · avg 45%	▇ 2.8%	sharp · at its level
Decision Index 0.4.1: GPQA Diamond (accuracy)	37.2% ✎	▂▆▅▅▆█▆▃▁··▁ beats 95/246	medium · avg 39%	▇ 2.5%	blunt · at its level

6 more tests

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: MuSR (accuracy)	59.0% ✎	▁▁▂█▅▄▁▁▁▁▁▁ beats 123/246	medium · avg 60%	▇ 1.1%	blunt · at its level
Decision Index 0.4.1: CRUXEval (accuracy)	87.9% ✎	██▄▃▃▄▁▁···▁ beats 232/246	medium · avg 58%	▇ 1.0%	at its level · noisy
Decision Index 0.4.1: SATA-Bench (case exact accuracy)	29.6% ✎	▁▁▁▁▁▁▁▂▅█▂▁ beats 138/246	hard · avg 22%	▇ 0.9%	typical
Decision Index 0.4.1: CLadder (accuracy)	85.3% ✎	▁███▆▅▃▃▁▁▁▁ beats 203/246	easy · avg 71%	▇ 0.9%	at its level
Decision Index 0.4.1: ChessBench (accuracy)	19.9% ✎	▄▇▄▄▄▆▄█▄▁▁▁ beats 169/246	hard · avg 16%	▇ 0.5%	blunt
Decision Index 0.4.1: HLE (accuracy)	9.0% ✎	▁▁▃▃▆█▇▅▅▃▁▁ beats 11/246	very hard · avg 13%	▇ 0.0%	blunt

other tests2 not in a task · 6.9% of its score

test	result	vs the others in its run	difficulty	counts	why
Decision Index 0.4.1: private same skills (equated points)	38.0900 points ✎	▂▂▂▃▃▃▄▅▅▇█▃ beats 88/247	avg 41.206 points	▇▇ 4.5%	blunt
Decision Index 0.4.1: private new domains (equated points)	27.8800 points ✎	▃█▃▃▂▆▃▃▄▄▄▄ beats 94/247	avg 38.861 points	▇ 2.5%	blunt

$ where to run it

Download model files ↗
Published model files verified. This is a verified repository snapshot; the benchmark does not establish an identical weight revision. A local run command and hardware requirements have not been verified for this entry. Source · checked 2026-10-31.

Model repository and documentation: huggingface.co/kpiya ↗

$ 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	42	▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 99%	CC BY 4.0
prior · pull toward the middle		▇ 1%
weighted average		= 46.1

not counted (2)

Decision Index 0.4.1	Decision Index 0.4.1: RouterBench (selected quality (quality objective)) [correct]	69.2%	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]	67.5%	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