we're expanding the http://
Character.AI universe and looking builders who dream big! searching for world-class builders, designers & researchers who want to make imagination interactive. if YOU or anyone you know in Product & Design, Engineering, and ML might be
MACHINE LEARNING
-
Character.AI Hiring World-Class Builders Designers Researchers
By
–
-
Demystifying AI Terminology: A Visual Guide to Key Concepts
By
–
AI is full of buzzwords—but what do they really mean? This visual breaks down the essentials: AI ML Deep Learning NLP Computer Vision Reinforcement Learning A clear map to cut through the jargon.
Which one do you think is most misunderstood? #AI -
Google DeepMind Dreamer 4 Mines Minecraft Using World Models
By
–
Crazy paper from Google DeepMind
— alphaXiv (@askalphaxiv) 7 octobre 2025
Dreamer 4 can mine Minecraft diamonds without ever touching the real game
It trained entirely inside its own world model, kinda like an “imaginary world”, showing agents can tackle long horizon tasks safely by just learning from videos! pic.twitter.com/5cG9oYyt51Crazy paper from Google DeepMind Dreamer 4 can mine Minecraft diamonds without ever touching the real game It trained entirely inside its own world model, kinda like an “imaginary world”, showing agents can tackle long horizon tasks safely by just learning from videos!
-
Scaling TorchTitan with SkyPilot Community Tutorials
By
–
More ways to scale TorchTitan with SkyPilot great to see the community expanding the ecosystem with new tutorials like this one!
-
VideoRAG: RAG Framework for Long-Context Video Comprehension
By
–
📢VideoRAG: Redefining Long-Context Video Comprehension
— Satya Mallick (@LearnOpenCV) 7 octobre 2025
In this week’s deep dive, we explore another interesting approach for performing RAG on videos. VideoRAG, a groundbreaking framework that brings RAG to the world of extremely long videos. Unlike traditional LVLMs that… pic.twitter.com/38rBOOlJjsRAG: Redefining Long-Context Comprehension In this week’s deep dive, we explore another interesting approach for performing RAG on videos. VideoRAG, a groundbreaking framework that brings RAG to the world of extremely long videos. Unlike traditional LVLMs that
-
DPT-2 Document Pre-Trained Transformer Model Revolutionizes Complex Table Parsing
By
–
The new Document Pre-Trained Transformer (DPT-2) model in ADE is a game changer. It parses complex tables with merged cells, multi-level headers, and irregular grids, with spatial grounding and bounding boxes for traceability. GitHub Repo:
-
SAS AI Machine Learning Skills Fest: Free Virtual Event November
By
–
Take your analytics skills further with the SAS AI and Machine Learning Skills Fest. 3 days of live sessions on ML, visual analytics, ethics in AI & more Free, virtual event with expert Q&A Nov. 12–14 | Starts 10 a.m. ET Save your seat: http://
2.sas.com/6013AiR6c -
Machine Learning for Healthcare 2025 Proceedings Published on PMLR
By
–
Volume 298 https://
proceedings.mlr.press/v298/ Proceedings of Machine Learning for Healthcare 2025 Is now available on PMLR -
Evolution Strategies at Scale Outperforms PPO for LLM Fine-Tuning
By
–
Evolution Strategies can be applied at scale to fine-tune LLMs, and outperforms PPO and GRPO in many model settings!
— hardmaru (@hardmaru) 7 octobre 2025
Fantastic paper “Evolution Strategies at Scale: LLM Fine-Tuning Beyond Reinforcement Learning” by @yule_gan, Risto Miikkulainen and team.https://t.co/CEyX6Z5ulG https://t.co/mPkXABJcQzEvolution Strategies can be applied at scale to fine-tune LLMs, and outperforms PPO and GRPO in many model settings! Fantastic paper “Evolution Strategies at Scale: LLM Fine-Tuning Beyond Reinforcement Learning” by @yule_gan
, Risto Miikkulainen and team. https://
arxiv.org/abs/2509.24372 -
Proprietary Data Essential for Meaningful Model Evaluation Beyond Public Benchmarks
By
–
I’d say using publicly available benchmarks and data are basically just sanity checks. As I mentioned mentioned in the conclusion, ideally you want/need to set up the eval with proprietary data that is related to your business problem, to make sure the performance is not due to