Goblin
News
AI news by
promptgoblins.ai
|
News
About
News
About
Filtered by:
Long Context
Clear
Titles
Summaries
6
Research: Four Minor Transformer Design Choices Can Tank Long-Context Performance by Up to 47%
Research
1
Aug 14
6
Research: Four Minor Transformer Design Choices Can Tank Long-Context Performance by Up to 47%
Research
· 1 src · Aug 14
Discuss
6
inclusionAI / Ant Group Releases Ling-3.0-Flash: Hybrid-Linear MoE with 256K Context and 60–80% TTFT Reduction
Updated Aug 4
Models
2
Aug 4
6
inclusionAI / Ant Group Releases Ling-3.0-Flash: Hybrid-Linear MoE with 256K Context and 60–80% TTFT Reduction
Upd Aug 4
Models
· 2 srcs · Aug 4
Discuss
6
Upstage Releases Solar Open 2: 250B Open-Weight MoE with 1M-Token Context and 15B Active Parameters
Models
1
Jul 22
6
Upstage Releases Solar Open 2: 250B Open-Weight MoE with 1M-Token Context and 15B Active Parameters
Models
· 1 src · Jul 22
Discuss
6
MiniMax Open-Sources Sparse Attention Library for Million-Token Contexts on NVIDIA Blackwell
Open Source
1
Jun 15
6
MiniMax Open-Sources Sparse Attention Library for Million-Token Contexts on NVIDIA Blackwell
Open Source
· 1 src · Jun 15
Discuss
Filters
Signal
Title
Category
Sources
Posted
Discuss
