ML applications

Perplexity AI Open-Sources Unigram Tokenizer That Achieves 5x Lower p50 Latency Than Hugging Face tokenizers Crate

ML applications

11 Min Read

Perplexity AI Open-Sources Unigram Tokenizer That Achieves 5x Lower p50 Latency Than Hugging Face tokenizers Crate

By Editorial Team

May 28, 2026

Read Article

Sakana AI Proposes DiffusionBlocks: a Block-wise Training Framework That Converts Residual Networks into Independently Trainable Denoising Modules

ML applications

11 Min Read

Sakana AI Proposes DiffusionBlocks: a Block-wise Training Framework That Converts Residual Networks into Independently Trainable Denoising Modules

By Editorial Team

May 28, 2026

Read Article

NVIDIA Releases Polar, a Token-Faithful Rollout Framework for GRPO Training Across Codex, Claude Code, and Qwen Code

ML applications

10 Min Read

NVIDIA Releases Polar, a Token-Faithful Rollout Framework for GRPO Training Across Codex, Claude Code, and Qwen Code

By Editorial Team

May 27, 2026

Read Article

Together AI Open-Sources OSCAR: An Attention-Aware 2-Bit KV Cache Quantization System for Long-Context LLM Serving

ML applications

10 Min Read

Together AI Open-Sources OSCAR: An Attention-Aware 2-Bit KV Cache Quantization System for Long-Context LLM Serving

By Editorial Team

May 26, 2026

Read Article

NVIDIA AI Releases Gated DeltaNet-2: A Linear Attention Layer That Decouples Erase and Write in the Delta Rule

ML applications

9 Min Read

NVIDIA AI Releases Gated DeltaNet-2: A Linear Attention Layer That Decouples Erase and Write in the Delta Rule

By Editorial Team

May 24, 2026

Read Article

Perplexity Open-Sources Bumblebee: A Read-Only Supply-Chain Scanner for Developer Endpoints

ML applications

5 Min Read

Perplexity Open-Sources Bumblebee: A Read-Only Supply-Chain Scanner for Developer Endpoints

By Editorial Team

May 23, 2026

Read Article

Nous Research Releases Contrastive Neuron Attribution (CNA): Sparse MLP Circuit Steering Without SAE Training or Weight Modification

ML applications

10 Min Read

Nous Research Releases Contrastive Neuron Attribution (CNA): Sparse MLP Circuit Steering Without SAE Training or Weight Modification

By Editorial Team

May 23, 2026

Read Article

Cohere Releases Command A+: A 218B Sparse MoE Model for Agentic Workflows That Runs on as Few as Two H100 GPUs

ML applications

5 Min Read

Cohere Releases Command A+: A 218B Sparse MoE Model for Agentic Workflows That Runs on as Few as Two H100 GPUs

By Editorial Team

May 22, 2026

Read Article

Microsoft Releases Fara1.5: A Family of Browser Computer-Use Agents (4B/9B/27B) That Outperform OpenAI Operator and Gemini 2.5 Computer Use on Online-Mind2Web

ML applications

5 Min Read

Microsoft Releases Fara1.5: A Family of Browser Computer-Use Agents (4B/9B/27B) That Outperform OpenAI Operator and Gemini 2.5 Computer Use on Online-Mind2Web

By Editorial Team

May 22, 2026

Read Article

Qwen Introduces Qwen3.7-Max: A Reasoning Agent Model With a 1M-Token Context Window

ML applications

9 Min Read

Qwen Introduces Qwen3.7-Max: A Reasoning Agent Model With a 1M-Token Context Window

By Editorial Team

May 22, 2026

Read Article

Popular Posts

US investors will soon get access to SK Hynix, another memory maker riding the AI

Training Gemma-3 for Structured Mathematical Reasoning with Tunix GRPO, LoRA Adapters, and GSM8K Rewards

Microsoft lays off nearly 5,000 employees across Xbox, commercial sales

Synthetic Sciences Releases OpenScience: An Open-Source, Model-Agnostic AI Workbench for Machine Learning, Biology, Physics, and

ML applications

Perplexity AI Open-Sources Unigram Tokenizer That Achieves 5x Lower p50 Latency Than Hugging Face tokenizers Crate

Sakana AI Proposes DiffusionBlocks: a Block-wise Training Framework That Converts Residual Networks into Independently Trainable Denoising Modules

NVIDIA Releases Polar, a Token-Faithful Rollout Framework for GRPO Training Across Codex, Claude Code, and Qwen Code

Together AI Open-Sources OSCAR: An Attention-Aware 2-Bit KV Cache Quantization System for Long-Context LLM Serving

NVIDIA AI Releases Gated DeltaNet-2: A Linear Attention Layer That Decouples Erase and Write in the Delta Rule

Perplexity Open-Sources Bumblebee: A Read-Only Supply-Chain Scanner for Developer Endpoints

Nous Research Releases Contrastive Neuron Attribution (CNA): Sparse MLP Circuit Steering Without SAE Training or Weight Modification

Cohere Releases Command A+: A 218B Sparse MoE Model for Agentic Workflows That Runs on as Few as Two H100 GPUs

Microsoft Releases Fara1.5: A Family of Browser Computer-Use Agents (4B/9B/27B) That Outperform OpenAI Operator and Gemini 2.5 Computer Use on Online-Mind2Web

Qwen Introduces Qwen3.7-Max: A Reasoning Agent Model With a 1M-Token Context Window

Recent Posts

Archives

Categories