Goblin
News
AI news by
promptgoblins.ai
|
News
About
News
About
Filtered by:
Reasoning Models
Clear
Titles
Summaries
6
BDH-CQ: 150M-Parameter Model Sets New ARC-AGI Cost-Efficiency Record at $0.0007 Per Task
Research
1
Aug 14
6
BDH-CQ: 150M-Parameter Model Sets New ARC-AGI Cost-Efficiency Record at $0.0007 Per Task
Research
· 1 src · Aug 14
Discuss
8
Microsoft Launches MAI-Thinking-1: Medium-Sized Reasoning Model Matches Claude Opus 4.6 on Coding Benchmarks
Models
1
Aug 13
8
Microsoft Launches MAI-Thinking-1: Medium-Sized Reasoning Model Matches Claude Opus 4.6 on Coding Benchmarks
Top
Models
· 1 src · Aug 13
Discuss
8
The Atlantic: AI Models From Four Major Labs Autonomously Hacked External Systems — "Time to Panic"
Security
1
Aug 12
8
The Atlantic: AI Models From Four Major Labs Autonomously Hacked External Systems — "Time to Panic"
Top
Security
· 1 src · Aug 12
Discuss
6
inclusionAI Releases Ling-3.0-tiny: 8B MoE Model Activating Only 1.3B Parameters for Local Deployment
Research
1
Aug 10
6
inclusionAI Releases Ling-3.0-tiny: 8B MoE Model Activating Only 1.3B Parameters for Local Deployment
Research
· 1 src · Aug 10
Discuss
9
OpenAI's AI Models Solve Multiple Erdős Mathematical Conjectures in Historic Breakthrough
Research
1
Aug 5
9
OpenAI's AI Models Solve Multiple Erdős Mathematical Conjectures in Historic Breakthrough
Top
Research
· 1 src · Aug 5
Discuss
7
Thinking Machines Releases Inkling-Small: Efficient 276B MoE Open-Weights Model That Outperforms Its Larger Predecessor on Reasoning
Products
1
Jul 31
7
Thinking Machines Releases Inkling-Small: Efficient 276B MoE Open-Weights Model That Outperforms Its Larger Predecessor on Reasoning
Products
· 1 src · Jul 31
Discuss
6
AI Reasoning Models Achieve IMO Gold Medals but Researchers Find Evidence of Shortcut-Taking Over Genuine Logic
Research
1
Jul 31
6
AI Reasoning Models Achieve IMO Gold Medals but Researchers Find Evidence of Shortcut-Taking Over Genuine Logic
Research
· 1 src · Jul 31
Discuss
8
AI Disproves Two Mathematical Conjectures in One Week, Signaling New "Brute Intelligence" Era
Research
1
Jul 27
8
AI Disproves Two Mathematical Conjectures in One Week, Signaling New "Brute Intelligence" Era
Top
Research
· 1 src · Jul 27
Discuss
6
'Antidoom' Technique Cuts LLM Reasoning Model Loop Failures by 7× Using Final Token Preference Optimization
Research
1
Jul 8
6
'Antidoom' Technique Cuts LLM Reasoning Model Loop Failures by 7× Using Final Token Preference Optimization
Research
· 1 src · Jul 8
Discuss
Filters
Signal
Title
Category
Sources
Posted
Discuss
