Buy Crypto
Pay with
USD
USD
Buy & Sell
Hot
Supports Visa, Mastercard, SEPA & more
P2P
0 Fees
Flexible trading, zero fees
Gate Card
Use your crypto for payments worldwide
Markets
Trade
Basic
Advanced
Futures
Futures
Access hundreds of perpetual contracts
CFD
Gold
One platform for global traditional assets
Options
Hot
Trade European-style vanilla options
Unified Account
Maximize your capital efficiency
Demo Trading
Introduction to Futures Trading
Learn the basics of futures trading
Futures Events
Join events to earn rewards
Demo Trading
Use virtual funds to practice risk-free trading
U.S. Stocks
CFD
U.S. stock CFD derivatives
Futures
High leverage, 24/7 trading
Tokenized Stocks
Backed by real stock assets
GUSD
Mint GUSD for Treasury RWA yields
Earn
Launch
CandyDrop
tag
Collect candies to earn airdrops
Launchpool
Quick staking, earn potential new tokens
HODLer Airdrop
Hold GT and get massive airdrops for free
Pre-IPOs
Unlock full access to global stock IPOs
Alpha Points
Trade on-chain assets and earn airdrops
Futures Points
Earn futures points and claim airdrop rewards
Investment
Square
Square
Post, share, and explore crypto trends
Live
moments live
Live crypto market analysis
Chat
Chat with crypto traders
News
What is happening in crypto
flower_head
More
Promotions
AI
Others
TradFi
Rewards
Square
Latest
Hot
News
Profile

A 2-billion-parameter small-model surpasses NVIDIA’s flagship model. Shanghai’s “World Model” tops an internationally recognized authoritative leaderboard with a “direct entry” (“no special preparation” / no special handling).

MeNews
robot
Abstract generation in progress
ME AI News, from ZhiYuan Robot, the self-developed world model Genie Envisioner-Sim 2.0 (abbreviated as GE 2.0) has achieved excellent overall performance, ranking first on the global world model evaluation benchmark World Arena's "Perception and Action Response" leaderboard. This time, ZhiYuan's GE 2.0 participated in the "Perception and Action Response" track evaluation, directly competing with top domestic and international AI teams such as NVIDIA's latest model DreamDojo and the Ctrl-World team jointly developed by Tsinghua University and Stanford, ultimately winning the championship. According to disclosed technical documents, GE 2.0, a model with only 2 billion (2B) parameters, outperformed flagship models with ultra-large parameters from NVIDIA, Microsoft, and others, also demonstrating that lightweight models are not inferior to ultra-large parameter models in humanoid robot applications. (Shanghai Observer News) (Source: Tonghuashun)
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • 7
  • Repost
  • Share
Comment
Add a comment
Add a comment
FiveMinutesBeforeLiquidation
· 5h ago
NVIDIA has actually been overtaken by 2B parameters; maybe it's time to rethink the parameter arms race.
View OriginalReply0
FeeswitchWhisperer
· 9h ago
Winning with a lightweight model shows that algorithm architecture matters far more than brute-force parameter stacking.
View OriginalReply0
NotificationSoundInMistyValley
· 9h ago
2B vs tens of billions, this efficiency gap is a matter of life and death on the edge side
View OriginalReply0
Re-StakingSucculents
· 10h ago
The key to humanoid robots hitting the ground is the world model; this track will become increasingly competitive.
View OriginalReply0
SudoSoul
· 10h ago
Perception and Action Response Track, in simple terms, is the robot version of "see and act."
View OriginalReply0
StakingDaydreamer
· 10h ago
Is the Ctrl-World team the one that came out of MIT—getting shredded by the 2B parameter?
View OriginalReply0
MountainBeforeTheStorm
· 10h ago
Have you heard of the World Arena benchmark before? Does anyone knowledgeable want to give a quick explanation?
View OriginalReply0

  • Pinned

Sitemap