NeuralSynthesis-7B-v0.1 Benchmark Results

NeuralSynthesis-7B-v0.1 — benchmark results

Kukedlc's model-stock merge of four Mistral-7B variants on a YamshadowExperiment28 base, built with LazyMergekit. Provider: Other. Released 2024-04-06. Access: Open.

Unified ELO 1404 ± 14, rank #1204 of 1776 rated models, from 15 benchmark results.

Strongest benchmark results

Benchmark	Score	Metric	Percentile
Open Chinese LLM - TruthfulQA MC	63.77	Accuracy (%)	95
Open LLM Leaderboard - MuSR	13.16	Score	72.4
Open Korean LLM Leaderboard	345.89	Average Score (%)	70.4
Open Chinese LLM - ARC Challenge	53.75	Accuracy (%)	59.8
Open Chinese LLM - HellaSwag	60.85	Accuracy (%)	58.8
Open LLM Leaderboard - BBH	31.83	Score	58.8
Open Chinese LLM - GSM8K	46.7	Accuracy (%)	54.5
Open Chinese LLM - WinoGrande	63.14	Accuracy (%)	53.6
Open Chinese LLM Leaderboard	56.74	Average Score (%)	53
Open LLM Leaderboard - IFEval	41.85	Score	42.6
Open LLM Leaderboard - MMLU-Pro	22.77	Score	38.1
Open LLM Leaderboard - GPQA	4.14	Score	36

Interactive version: theaggregate.ai/model?slug=neuralsynthesis-7b-v0-1 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.