PulseAugur
中
实时 16:48:06
English(EN) Nous Research has released Arete-27B, an Apache 2.0 licensed fine-tune of Qwen3.8-27B aimed at cutting political bias out of model responses. On the company's o

Nous Research 微调 Qwen3.8-27B，将宣传内容减少 98%

作者 PulseAugur 编辑部 · [3 个来源] ·

Nous Research 发布了 Arete-27B，这是一个从阿里巴巴的 Qwen3.8-27B 微调而来的开源权重模型。该新模型显著减少了响应中的宣传内容，在其内部评估中从 22% 降至 0.9%。在保持 Hermes Bench 等基准测试相似性能的同时，Arete-27B 旨在提供 Qwen 级别的能力，但没有国家意识形态的框架。 AI

影响 提供了一种利用强大基础模型同时减轻政治偏见输出的方法。

排序理由 发布了一个具有特定性能改进的微调开源权重模型。

在 Mastodon — mastodon.social 阅读 →

AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →

Nous Research 微调 Qwen3.8-27B，将宣传内容减少 98%

本文如何被排名

Signal score
5 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Research
发布了一个具有特定性能改进的微调开源权重模型。
Source corroboration
3 independent sources
Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics
model release, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Same-day
Cluster formed today. Ranking reflects the current source set at time of score.
Coverage growth since scoring
+1 source(s) since last score
New sources have picked up this story since our last re-score. Score will update on the next scoring pass.

完整方法见我们的编辑标准。

报道来源 [3]

  1. MarkTechPost TIER_1 English(EN) · Asif Razzaq ·

    Nous Research发布Arete-27B，将Qwen3.8-27B的宣传内容占比从22%降至0.9%

    <p>Nous Research released Arete-27B, an Apache 2.0 fine-tune of Qwen3.8-27B. On Nous’s 803-prompt eval, propaganda answers fell from 22% to 0.9% while Hermes Bench stayed flat at 65.60%.</p> <p>The post <a href="https://www.marktechpost.com/2026/11/07/nous-research-arete-27b-qwen…

  2. Mastodon — mastodon.social TIER_1 English(EN) · Steenbergen_apps ·

    📰 Latent — 一个 27B 模型去宣传化，加上一个 vLLM 控制平面 • Nous Research 发布 Arete-27B，一个几乎消除宣传的 Qwen3.8 微调版本

    📰 Latent — A 27B model de-propagandised, plus a vLLM control plane • Nous Research releases Arete-27B, a Qwen3.8 fine-tune that nearly eliminates propaganda answers • Buoyant open-sources Kinference, an inference control plane for vLLM on Kubernetes • The US cleared Nvidia H200 s…

    链接 latent.steenbergen.ai/…/2026-11-08 latent.steenbergen.ai/archive
  3. Mastodon — mastodon.social TIER_1 English(EN) · [email protected] ·

    Nous Research 发布了 Arete-27B，一个 Apache 2.0 许可的 Qwen3.8-27B 微调模型，旨在消除模型响应中的政治偏见。在该公司官

    Nous Research has released Arete-27B, an Apache 2.0 licensed fine-tune of Qwen3.8-27B aimed at cutting political bias out of model responses. On the company's own 803-prompt evaluation, answers classified as propaganda fell from 22% to 0.9%, while scores on the Hermes Bench staye…

    链接 marktechpost.com/…/nous-research-arete-27…

相关实体

相关话题