whatmodel.sh / decision / Kev-4B (JevBench mini)
Kev-4B (JevBench mini)
↗ share
+ contribute
Kev-4B (JevBench mini) is a source-available decision implementation from Jared Palmer.
Catalogue identity notes
JevBench 0.1.0 mini tested configuration kev-4b: Decoder, pointer head; base Qwen3.5-4B-Base; readout rank-16 LoRA + pointer head. Separate identity because the report does not establish equivalence to the newer Decision Index checkpoint or serving recipe. No latency imported.
$ how it compares
The decision models ranked either side of Kev-4B (JevBench mini), out of 357 with a score, then more from Jared Palmer.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 238 |  OpenThai-SystemOne (JevBench mini) | 46±2 | source code | ||
| 239 |  vera-core-2 | 46±1 | model files | ||
| 240 | Kev-4B (JevBench mini) | 46±2 | source code | ||
| 241 |  TokenRhythm NeoHorse-Jev-4B | 46±1 | model files | ||
| 242 | SemIf Qwen3.5-4B Base (decision readout) | 46±1 | source code | ||
| more from Jared Palmer | |||||
| 88 | Kev 27B | 62±1 | model files | ||
| 159 | kev-4b (DecisionBench 485ace87) | 55±2 | model files | ||
| 162 | kev-8b (DecisionBench c80773da) | 55±2 | model files | ||
see all decision models · find the ones that fit your machine
$ what it's good at
consistency▃·▂▂·▅▃·▃▅▃█ #14 of 21 · 60% of its score · 5 tests
Trails RSI-Jev v3.0 2B (JevBench mini) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| JevBench mini 0.1.0 · Probability measure coherence (pass rate) | 30.1% ✎ | ▄·▁·▃▃▁·▃▄▁█ beats 7/20 | 43% |
| JevBench mini 0.1.0 · Representation consistency (pass rate) | 74.4% ✎ | ▃·▂·▂▂▂▃▅▅▃█ beats 7/20 | 14% |
| JevBench mini 0.1.0 · Logical coherence (pass rate) | 67.5% ✎ | ▃·▃·█▅▅██▃█▅ beats 3/20 | 2.3% |
| JevBench mini 0.1.0 · Choice-set coherence (pass rate) | 87.5% ✎ | ▁▃▄█▄▄·····▄ beats 15/20 | 1.2% |
| JevBench mini 0.1.0 · Batch independence (pass rate) | 100.0% ✎ | ▁··········█ beats 9/20 | 0.0% |
other tests1 not in a task · 38% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| JevBench mini 0.1.0 · unchanged-answer accuracy | 87.9% ✎ | ▁·▂·▂···▂▂▁█ beats 12/20 | 38% |
$ 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/jaredpalmer ↗
$ how the score was worked out · JevBench 0.1.0 mini evaluatio… 98%
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 |
|---|---|---|
 JevBench 0.1.0 mini evaluation reports | 6 | 98% |
| prior · pull toward the middle | 2% | |
| weighted average | = 45.8 |
$ improve this page
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 · sources and licences · json