PulseAugur
中
实时 09:52:07
English(EN) 📊 DeepSeek V4 Flash 0731 (Reasoning, Max Effort) scores 90.8% GPQA and 38.6% on HLE, with 233.8 tokens/sec. At 52.3 intelligence points per dollar, it's a stand

DeepSeek-V4 Flash 0731 在 GPQA 和 HLE 上取得强劲分数

作者 PulseAugur 编辑部 · [1 个来源] ·

DeepSeek-V4 Flash 0731 在 GPQA 基准测试中取得了 90.8% 的分数，在 HLE 上取得了 38.6% 的分数。该模型还展示了每秒 233.8 个 token 的速度。值得注意的是，它提供了高智能成本比，每美元提供 52.3 个智能点，在大型语言模型领域中成为一个高效的选择。 AI

影响 在大型语言模型基准测试中展示了竞争力的性能和效率。

排序理由 发布了模型基准测试结果。[lever_c_demoted from research: ic=1 ai=1.0]

在 Mastodon — fosstodon.org 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

DeepSeek-V4 Flash 0731 在 GPQA 和 HLE 上取得强劲分数

本文如何被排名

Signal score
10 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
发布了模型基准测试结果。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
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
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

  1. Mastodon — fosstodon.org TIER_1 English(EN) · [email protected] ·

    📊 DeepSeek V4 Flash 0731 (Reasoning, Max Effort) 在 GPQA 上得分 90.8%，在 HLE 上得分 38.6%，速度为 233.8 tokens/sec。每美元 52.3 智能点，它是一个领先

    📊 DeepSeek V4 Flash 0731 (Reasoning, Max Effort) scores 90.8% GPQA and 38.6% on HLE, with 233.8 tokens/sec. At 52.3 intelligence points per dollar, it's a standout for efficiency. Measured independently, not self-reported → https:// olud.ai/leaderboard.html # LLM # Benchmarks # O…

    链接 olud.ai/leaderboard.html

相关实体

相关话题