Large language models have shown remarkable potential in text generation, but video interpretation remains a challenge. How can we bridge this gap? Enter Video-STaR: a promising self-training approach to video.
MACHINE LEARNING
-
Call to open-source Grok 2 on a perfect occasion
By
–
Awesome news! Wouldn't that be a perfect occasion to open-source Grok 2!
-
TTRL Boosts LLM Reasoning by 159% Using Unlabeled Test Data
By
–
LLMs usually rely on labeled data to improve But what if they could self-improve at test time — without any labels? Introducing TTRL: a new RL framework that boosts LLM reasoning by up to 159% on the AIME 2024 using just unlabeled test data Trending #1 on alphaXiv
-
Scaling Laws For Scalable Oversight
By
–
Scaling Laws For Scalable Oversight
Link: https://
arxiv.org/pdf/2504.18530 -
Alibaba Qwen3: Eight Open Models with Advanced Thinking Capabilities
By
–
Alibaba’s Qwen just unveiled Qwen3: a family of eight open models ranging from 600M to 235B params.
— Rowan Cheung (@rowancheung) 29 avril 2025
— Flagship version rivals OpenAI o1 & DeepSeek-R1
— Hybrid “thinking” mode in all models
— Boosted coding + agent performance
— Supports 119 languagespic.twitter.com/eM5mw2TjIPAlibaba’s Qwen just unveiled Qwen3: a family of eight open models ranging from 600M to 235B params. — Flagship version rivals OpenAI o1 & DeepSeek-R1
— Hybrid “thinking” mode in all models
— Boosted coding + agent performance
— Supports 119 languages -
Google Demos Visual Blocks and InstructPipe ML Prototyping Tools
By
–
Stop by the #CHI2025 Google Booth at 3:40PM today to catch a demo of Visual Blocks, a platform for rapid prototyping of ML-based multimedia applications, & InstructPipe, a Gemini-based AI assistant for prototyping ML pipelines with text instructions. https://
visualblocks.withgoogle.com -
Large-Scale AGI Business Demo: Infinite Bloom 3.0
By
–
Welcome to the era of "Large‑Scale α‑AGI Businesses". Large‑Scale α‑AGI Business ($AGIALPHA) Demo – “Infinite Bloom 3.0”: https://
github.com/MontrealAI/AGI
-Alpha-Agent-v0/tree/main/alpha_factory_v1/demos/alpha_agi_business_2_v1
… #AIAgent #AGI #AGIALPHA -
TensorFlow.js Implementation with Canvas Training Visualization
By
–
"Write an implementation of this paper using Tensorflow.js and visualize the training with Canvas."
— Jeff Dean (@JeffDean) 28 avril 2025
Thanks, Gemini! https://t.co/718fKWV25V"Write an implementation of this paper using Tensorflow.js and visualize the training with Canvas." Thanks, Gemini!
-
Qwen 3 : 6 nouveaux modèles, dont le SOTA 235B
By
–
BREAKING : Qwen released 6 new Qwen 3 family models, including SOTA "Qwen3-235B-A22B" with top-tier performance. Models are accompanied by better support for coding and MCP tasks.
-
How Coders Use AI: Economic Index Insights
By
–
New from the Anthropic Economic Index: Evidence on how coders use AI. Our previous research confirmed that AI is disproportionately used for software development work. Here, we go into the details.