Looks like DeepSeek is temporarily down. @deepseek_ai
@jiqizhixin
-
OpenAI Google DeepMind DeepSeek launch AI models same day
By
–
Woke up to find OpenAI launching 4o Image Generation and Google DeepMind dropping Gemini 2.5 at the same time. What a day! Especially knowing that DeepSeek-V3 and Qwen2.5-VL-32B-Instruct just landed yesterday!
-
Video-T1: Test-Time Scaling for Video Generation
By
–
-T1: Test-Time Scaling for Generation
Paper: https://
arxiv.org/pdf/2503.18942
Page: https://
liuff19.github.io/Video-T1/
Code: -
Test-Time Compute Improves Video Quality Significantly
By
–
And what did they find? Increasing test-time compute consistently leads to significant improvements in video quality.
-
Video Generation Models Leverage Inference Compute for Quality
By
–
If a video generation model is allowed to use a non-trivial amount of inference-time compute, how much can it improve generation quality for challenging text prompts?
This is exactly what a new paper from Tsinghua University explores. -
ARC-AGI-2 Benchmark Shows AI Reasoning at 4% Accuracy
By
–
On the newly launched ARC-AGI-2 benchmark, the best reasoning AI today can only achieve 4% accuracy. https://
arcprize.org/leaderboard -
Qwen2.5-VL-32B-Instruct Released with RL Optimizations
By
–
Qwen2.5-VL-32B-Instruct is here! Optimized with RL, it shows significant improvements in human preference and mathematical reasoning.
-
DeepSeek V3 Model Update Released on Hugging Face
By
–
DeepSeek just rolled out an update for its V3 model! DeepSeek-V3-0324
Hugging Face: https://
huggingface.co/deepseek-ai/De
epSeek-V3-0324/tree/main
… -
Can You Really Be a Vibe Coder With Three Years Experience?
By
–
Can you really be a "vibe coder" with three years of experience when the term was just coined days ago?
-
Scientific Article and Source Code Shared on arXiv
By
–
Paper: https://arxiv.org/pdf/2503.05383
Code: