加载中...

CL

Claude3-Sonnet

Name: Claude3-Sonnet
Author: Anthropic

Claude3-Sonnet

Release date: 2024-03-04更新于: 2024-04-18 08:32:37669

Live demo

GitHub

Hugging Face

Compare

Parameters

Not disclosed

Context length

200K

Chinese support

Supported

Reasoning ability

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Claude3-Sonnet

Model basics

Reasoning traces

Not supported

Thinking modes

Thinking modes not supported

Context length

200K tokens

Max output length

No data

Model type

多模态大模型

Release date

2024-03-04

Model file size

No data

MoE architecture

No

Total params / Active params

0.0B / N/A

Knowledge cutoff

No data

Claude3-Sonnet

Open source & experience

Code license

不开源

Weights license

不开源- 不开源

GitHub repo

GitHub link unavailable

Hugging Face

Hugging Face link unavailable

Live demo

No live demo

Claude3-Sonnet

Official resources

Paper

Introducing the next generation of Claude

DataLearnerAI blog

评测结果超过GPT-4，Anthropic发布第三代大语言模型Claude3，具有多模态能力，实际评测表现优秀！

Claude3-Sonnet

API details

API speed

No data

No public API pricing yet.

Claude3-Sonnet

Benchmark Results

Claude3-Sonnet currently shows benchmark results led by MMLU Pro (97 / 114, score 56.80). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

综合评估

1 evaluations

Benchmark / mode

Score

Rank/total

MMLU Pro

Off

56.80

97 / 114

View benchmark analysis Compare with other models

Claude3-Sonnet

Publisher

Anthropic

View publisher details

Claude3-Sonnet

Model Overview

Claude3-Sonnet是Anthropic公司发布的第三代多模态大语言模型。第三代的Claude-3模型包含3个版本，这里说的Claude3-Sonnet是其中能力与速度最均衡的模型。各项评测人任务结果都非常好，能力介于Claude3-Opus和Claude3-Haiku之间。Claude3-Sonnet本身的速度是Claude2的2倍左右，但是各项测评结果都远优于Claude2！

在多模态方面，Claude3-Sonnet也有强大的能力。

Claude2最受诟病的就是无效的拒绝回答。由于Anthropic在对齐方面做了严格的工作，导致Claude2.1经常出现拒绝回答的情况。在Claude3-Opus上。Anthropic做了改进，在内部测试中，Claude2.1错误地拒绝比例大概在26%左右，而Claude3-Sonnet上这个比例下降到了10%多一点点，进步明显！也好与Claude3-Opus。

DataLearner 官方微信

欢迎关注 DataLearner 官方微信，获得最新 AI 技术推送