This research was led by Jackson Kaunismaa through the MATS program and supervised by researchers at Anthropic, with additional support from Surge AI and Scale AI. Read the full paper: https://
arxiv.org/pdf/2601.13528
INNOVATION
-
Anthropic-Led Research on AI Model Improvement Published
By
–
-

7-Eleven Deploys AI Agent to Accelerate Store Equipment Maintenance
By
–
.
@7eleven store maintenance technicians keep stores running smoothly by servicing a wide range of equipment — from food service appliances and refrigeration units to fuel dispensers and Slurpee machines. To support faster fix times, 7-Eleven built an AI agent that provides -
High-Sparsity Attention Tuning for Video Diffusion Transformers
By
–
SALAD Achieve High-Sparsity Attention via Efficient Linear Attention Tuning for Diffusion Transformer
-
Jet-RL: On-Policy FP8 Reinforcement Learning with Unified Precision
By
–
Jet-RL Enabling On-Policy FP8 Reinforcement Learning with Unified Training and Rollout Precision Flow
-
MCP Apps: Interactive Interfaces for Tools in Claude
By
–
We’ve launched the first official extension to MCP.
— Alex Albert (@alexalbert__) 26 janvier 2026
MCP Apps lets tools return interactive interfaces instead of just plain text.
Live in Claude today across a range of tools. https://t.co/aw3G3yxQvIWe’ve launched the first official extension to MCP. MCP Apps lets tools return interactive interfaces instead of just plain text. Live in Claude today across a range of tools.
-
Israel Meteorological Service Deploys NVIDIA Earth-2 AI Weather Model
By
–
To streamline accurate weather forecasting, read how Israel Meteorological Service (IMS) implemented an NVIDIA Earth-2 AI weather model into its operational weather forecasting system. https://
nvda.ws/4k3MDwq #AMS2026 @amirgiv @METEOROLOGY_IL -

Meta Unveils Maia 200 Inference Chip with Superior Performance
By
–
Our Maia 200 inference chip, announced today, is most performant first party silicon of any hyperscaler. 3x the FP4 performance of the Amazon Trainium v3, and FP8 performance above Google’s TPUv7.
-
Machines of Loving Grace: AI potential when implemented correctly
By
–
It's a companion to Machines of Loving Grace, an essay I wrote over a year ago, which focused on what powerful AI could achieve if we get it right:
-
GitHub Integration for Codebase Edits with Mobile Dictation
By
–
Yes but it’s a GH integration for codebase edits (ie not what I use Claude code for) but if that’s what you need, then you’ll have a mobile experience with dictation capabilities!
-
Superintelligence Team First to Use Maia 200 Hardware
By
–
It's a big day. Our Superintelligence team will be the first to use Maia 200 as we develop our frontier AI models. https://t.co/3Vt38ajobR
— Mustafa Suleyman (@mustafasuleyman) 26 janvier 2026It's a big day. Our Superintelligence team will be the first to use Maia 200 as we develop our frontier AI models.