Our AI+Science group from @Caltech has a number of papers at @NeurIPSConf in chronological order: FunDPS: Guided Diffusion Sampling on Function Spaces with Applications to PDEs https://
arxiv.org/abs/2505.17004
R-KV: Redundancy-aware KV Cache Compression for Reasoning Models
INNOVATION
-
Caltech AI+Science Group Presents NeurIPS Papers on Diffusion and LLMs
By
–
-

Kling 2.6 Launches with Enhanced Photorealism and Audio Generation
By
–
Kling 2.6 is now in Krea.
— KREA AI (@krea_ai) 3 décembre 2025
this new version comes with improved photorealism and audio generation.
try it now! pic.twitter.com/0aG6GRGq4kKling 2.6 is now in Krea. this new version comes with improved photorealism and audio generation. try it now!
-
AI Engineer Fixes Apps Like a Human
By
–
BREAKING: This AI software engineer uses your App like a human & fixes it.
— God of Prompt (@godofprompt) 3 décembre 2025
Lovable, Bolt, Cursor can't do this yet.
Anything max tests its own work. like a real engineer would.
700,000+ builders are already using Anything Max, the first autonomous software engineer that… pic.twitter.com/LJmkbMn0vJBREAKING: This AI software engineer uses your App like a human & fixes it. Lovable, Bolt, Cursor can't do this yet. Anything max tests its own work. like a real engineer would. 700,000+ builders are already using Anything Max, the first autonomous software engineer that
-

MIT Study Shows Ineffective Neural Networks Need Better Starting Points
By
–
Do "untrainable" neural networks just need a helping hand to do modern tasks? MIT study reveals many "ineffective" networks may start from suboptimal starting points. CSAIL's short-term guidance method transferred structural knowledge to boost performance:
-
Veo 3.1 vs Veo 2: 10 months of video generation progress
By
–
Amazing progress in video generation, in 10 months:
— m_ric (@AymericRoucher) 3 décembre 2025
Veo 3.1 (Oct '25) || Veo 2 (Dec '25) pic.twitter.com/SGaRWJ366gAmazing progress in video generation, in 10 months: Veo 3.1 (Oct '25) || Veo 2 (Dec '25)
-
Kling 2.6 Announces Native Audio and Quality Improvements
By
–
🔴 ¡KLING ANUNCIA KLING 2.6!
— Carlos Santana (@DotCSV) 3 décembre 2025
Continúa la semana de los anuncios en el panorama de la generación de vídeo, esta vez con una nueva versión de Kling 2.6, que mejora calidad y además incorpora audio nativo, similar a Veo 3 o Sora.
La verdad, muy buen primer resultado! pic.twitter.com/2iNYIgMbAd¡KLING ANUNCIA KLING 2.6! Continúa la semana de los anuncios en el panorama de la generación de vídeo, esta vez con una nueva versión de Kling 2.6, que mejora calidad y además incorpora audio nativo, similar a Veo 3 o Sora. La verdad, muy buen primer resultado!
-

MIT Mathematics System Ensures Safe Flexible Robot Interactions
By
–
MIT system uses rigorous mathematics to ensure robots flex, adapt, & interact w/people & objects in a safe & precise way. It helps robots remain flexible & responsive w/o exceeding safe force limits: https://
bit.ly/3KynwV0 -

SAS Hackathon Awards Announce Champion Winners December 11
By
–
More than 90 teams crossed the finish line And all finishers are winners Still, we can't wait to reveal which teams will solidify champion bragging rights… Tune in for our live #SASHackathon awards event on December 11 to find out! http://
2.sas.com/61817iFDH -
Curiosity Cup 2026: Global Analytics Competition for Students
By
–
Students, it’s time! Registration is now open for the 2026 Curiosity Cup. Take your analytics skills to the next level, compete with students worldwide and unlock incredible opportunities. Register your team by Feb 1:
-
Building Coding Agents with Tool Execution Course
By
–
New course: Building Coding Agents with Tool Execution, taught by @tereza_tizkova and @FraZuppichini from @e2b.
— Andrew Ng (@AndrewYNg) 3 décembre 2025
Most AI agents are limited to predefined function calls. This short course teaches you to build agents that write and execute code to accomplish tasks, accessing… pic.twitter.com/yUk3OH7MrLNew course: Building Coding Agents with Tool Execution, taught by @tereza_tizkova and @FraZuppichini from @e2b
. Most AI agents are limited to predefined function calls. This short course teaches you to build agents that write and execute code to accomplish tasks, accessing