Scale AI

@scale_AI

making AI work

scale.com

七月 2016 加入

2K帖子 72K关注者 482正在关注

你可能会喜欢

@alexandr_wang

@huggingface

@geoffreyhinton

@karpathy

@llama_index

@janleike

@AnthropicAI

@AIatMeta

@ClementDelangue

@LangChainAI

@a16z

@trychroma

@runwayml

@gdb

@ilyasut

Scale AI

年11月25日

👀

Igor Kotenkov

@stalkermustang

年11月24日

real metrics banger is hidden in the system card. Yes, you can overfit on Django and nail SWE-bench Verified. But there's this recent SWE-bench Pro from @scale_AI , and opus gets 52%. The next best, sonet 4.5, is only 43.6, and non-anthropic model, GPT-5, is 36%. This is HUGE…

stalkermustang's tweet image. real metrics banger is hidden in the system card. Yes, you can overfit on Django and nail SWE-bench Verified. But there's this recent SWE-bench Pro from @scale_AI , and opus gets 52%. The next best, sonet 4.5, is only 43.6, and non-anthropic model, GPT-5, is 36%.

This is HUGE…

stalkermustang's tweet image. real metrics banger is hidden in the system card. Yes, you can overfit on Django and nail SWE-bench Verified. But there's this recent SWE-bench Pro from @scale_AI , and opus gets 52%. The next best, sonet 4.5, is only 43.6, and non-anthropic model, GPT-5, is 36%.

This is HUGE…

Scale AI 已转帖

Udari Madhushani Sehwag

年11月21日

Our new research from @Scale_AI reveals that harmful biological knowledge can persist inside bio-foundation models even after filtering. We introduce BioRiskEval, the first comprehensive framework built to assess dual-use risk in these models using a realistic adversarial threat…

UdariMadhu's tweet image. Our new research from @Scale_AI reveals that harmful biological knowledge can persist inside bio-foundation models even after filtering.

We introduce BioRiskEval, the first comprehensive framework built to assess dual-use risk in these models using a realistic adversarial threat…

Scale AI 已转帖

Jeff Da

年11月13日

1/ New Chain of Thought podcast episode from @Scale_AI, where we dive into SWE-Bench Pro, a benchmark designed to rigorously evaluate LLM coding agents on professional software engineering tasks.

Scale AI 已转帖

Manasi Sharma

年11月12日

🚀New @scale_AI paper: 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵𝗥𝘂𝗯𝗿𝗶𝗰𝘀, a benchmark for evaluating Deep Research (DR) agents. Even top agents like Gemini & OpenAI DR achieve <𝟲𝟴% 𝗿𝘂𝗯𝗿𝗶𝗰 𝗰𝗼𝗺𝗽𝗹𝗶𝗮𝗻𝗰𝗲. We built 𝟮.𝟱𝗞+ expert rubrics with 𝟮.𝟴𝗞+ hrs of human labor to measure why.

ManasiSharma_'s tweet image. 🚀New @scale_AI paper: 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵𝗥𝘂𝗯𝗿𝗶𝗰𝘀, a benchmark for evaluating Deep Research (DR) agents. Even top agents like Gemini &amp; OpenAI DR achieve &lt;𝟲𝟴% 𝗿𝘂𝗯𝗿𝗶𝗰 𝗰𝗼𝗺𝗽𝗹𝗶𝗮𝗻𝗰𝗲. We built 𝟮.𝟱𝗞+ expert rubrics with 𝟮.𝟴𝗞+ hrs of human labor to measure why.

Scale AI

年11月11日

Today, we honor the courage, dedication, and sacrifice of all who have served. Thank you to our veterans. 🇺🇸

Scale AI

年11月10日

Scale 🤝 TIME Today, @TIME rolled out a Scale-powered, site-wide AI reading and discovery experience. The AI agent is just the latest activation in our ongoing partnership to help enhance access to journalism worldwide. Learn more via @axios: scl.ai/47BPIzm

scale_AI's tweet image. Scale 🤝 TIME

Today, @TIME rolled out a Scale-powered, site-wide AI reading and discovery experience.

The AI agent is just the latest activation in our ongoing partnership to help enhance access to journalism worldwide.

Learn more via @axios: scl.ai/47BPIzm

Alexandr Wang

@alexandr_wang

Riley Goodside

@goodside

hardmaru

@hardmaru

AK

@_akhaliq

Jim Fan

@DrJimFan

AI at Meta

@AIatMeta

AI Pub

@ai__pub

AI Breakfast

@AiBreakfast

Jay Hack

@mathemagic1an

Sharif Shameem

@sharifshameem

Emm | scenario.com

@emmanuel_2m

Eric Jang

@ericjang11

Tanishq Mathew Abraham, Ph.D.

@iScienceLuvr

Ksenia_TuringPost

@TheTuringPost

Jerry Liu

@jerryjliu0

Jack Clark

@jackclarkSF

Mo Bavarian

@mobav0

Gill Verdon

@GillVerd

Ramsri Goutham Golla

@ramsri_goutham

T

@tbddddddd

AyodhyaRam Mohanthy

@ayodhyamohanthy

JerseyHODL

@MikeyGio85

Briojuil

@Briojuil32835

Anjan_Mallik

@Krishna66148509

Naman Narang

@narangnaman29

Matías Nicolás Sosa

@sous_8

Gheorghe Cătălin Alexandru

@GheorgheCtlinA1

Maier Bianchi

@maierb

Степан Козлов

@stepanda200

Jarosz

@jarosz

Siyue Mousy

@MousySiyue

IndustrialX

@IndustrialXai

John Baptist

@johnbaptist42

RaeLambert

@p0h881bCCo3ktoI

Aakash Parikh

@imAParikh

Pasquale

@Hsushengchan

AirborneChao

@airbornechao

♈︎ Brooke Parker ♀

@BookBrookeXx

Sean Finn

@seanpfinn

Güler Ardıç

@gulerkovboy

Alonso Alarcón Aguilar

@AlonsoAlar12835

Huseyin Karayagiz

@huseyin_krygz

IKAI

@ikaiworld

Hataalii 🙏🏾🤴🏾✨

@jrmoyler

mithat

@temizkanmithat

@foundrceo

@foundrceo

x

@oO01iIlL

emine nur

@dogalolaraknur

Saul Berenson

@tryng2ndrstnd

Carrot 菜姐 @CHD

@Carrot_web3

Anmol Sinha

@553Anmol

wecoco

@wecochp

Han

@hanshut0ng

Justice Apex

@JusticeApexLLC

Billion

@billion_wallet

Rodney Lafuente-Mercado

@rodney_lafuente

KevinY

@creolophus123

Ali Tolga BAYINDIR

@alitolga82

mikey567

@bvi7823423

Brecht Waeghe

@_brecht

Chukwunonso Prosper

@prospercode

Shayan Shabihi

@shabihish

Clark Chen

@clarkchenkai

Hasan Baki Küçükçakıroğlu

@kucukcakiroglu_

Aero

@aerowindwalker

Juan Tovar

@JuanPTovar

Unseen Futures

@unseen_futures

Michael Gasvoda

@MichaelGasvoda

飞

@ovosuperfun

Alexandr Wang

@alexandr_wang

Riley Goodside

@goodside

Greg Brockman

@gdb

hardmaru

@hardmaru

AK

@_akhaliq

Emad

@EMostaque

Suhail

@Suhail

Google DeepMind

@GoogleDeepMind

Jim Fan

@DrJimFan

AI at Meta

@AIatMeta

Stability AI

@StabilityAI

Nat Friedman

@natfriedman

Nathan Benaich

@nathanbenaich

Runway

@runwayml

Jeremy Howard

@jeremyphoward

Soumith Chintala

@soumithchintala

Google AI

@GoogleAI

clem 🤗

@ClementDelangue

PyTorch

@PyTorch

Anthropic

@AnthropicAI

Chen Bo Calvin Zhang

@calvincbzhang

Afra Feyza Akyürek

@afeyzaakyurek

advait

@agxsai

Tom Channick

@tomchannick

Brad Kenstler

@Bckenstler

Chetan Rane

@the_chet2an

Bing Liu

@vbingliu

Matthew Siegel

@LargerLanguage

Jason Droege

@jdroege

Boris Power

@BorisMPower

Dan Hendrycks

@hendrycks

Center for AI Safety

@ai_risks

Lulu Cheng Meservey

@lulumeservey

Sean Hendryx

@SeanHendryx

lmarena.ai

@arena

Ben Lang

@benln

United Nations Institute for Disarmament Research

@UNIDIR

Teddy Schleifer

@teddyschleifer

Will Henshall

@henshall_will

Gilead Sciences

@GileadSciences

Mark Satter

@marksatter

Jonathan Lehrfeld

@lehrfeld_media

Oma Seddiq

@omaseddiq

National Institute of Standards and Technology

@NIST

U.S. Commerce Dept.

@CommerceGov

Gerard Baker

@gerardtbaker

Guy Taylor

@guyjtaylor

CSIS Defense and Security Department

@CSISDefense

Adrienne Watson

@Watson_Adrienne

Sean Savett Archived

@NSC_Spox46

Summer Yue

@summeryue0

Richard Nieva

@richardjnieva

Daniel Berrios

@danielxberrios

Benny (杜本立)

@bennydu

Washington Post Live

@PostLive

Chad Pergram

@ChadPergram

The OSS Society

@osssociety

Vijay Karunamurthy

@vjkaruna

ARCHIVED: Jen Easterly

@CISAJen

Benjamin Powers

@benjaminopowers

CNN Politics

@CNNPolitics

Techmeme

@Techmeme

Cat Zakrzewski

@Cat_Zakrzewski

MarketWatch

@MarketWatch

WydenPress

@WydenPress

Ron Wyden

@RonWyden

Washington AI Network

@WashingtonAINet

Tammy Haddad

@haddadmedia

Tom Lee-Devlin

@TomLeeDevlin

United States 趋势

1. Arch Manning 2,760 posts
2. Texas A&M 10.3K posts
3. #SmackDown 14.3K posts
4. #BedBathandBeyondisBack 1,410 posts
5. Sark 2,450 posts
6. Bears 127K posts
7. Aggies 5,243 posts
8. Eagles 141K posts
9. #OPLive 1,696 posts
10. #HookEm 4,210 posts
11. #iufb 2,401 posts
12. Marcel Reed 1,703 posts
13. Brunson 5,334 posts
14. Ben Johnson 27.3K posts
15. Wingo 1,763 posts
16. Lindor 1,896 posts
17. Purdue 5,491 posts
18. Bucks 18.7K posts
19. Longhorns 3,526 posts
20. Wisner N/A

你可能会喜欢

Something went wrong.

Something went wrong.