Cohere's North 2 Gives Enterprise AI Agents Persistent Memory and Tighter Controls
Cohere's North 2 platform adds persistent agent memory, 15+ new features, and connectors to Slack, GitHub, and Notion for sovereign enterprise AI deployments.
20 stories
Cohere's North 2 platform adds persistent agent memory, 15+ new features, and connectors to Slack, GitHub, and Notion for sovereign enterprise AI deployments.
A compact SigLIP2 fine-tune hits 97.2% accuracy on binary gender classification, pulling 778K downloads with a 93M-parameter footprint.
Stanford's Mulligan targets a robot's weak spots by restarting failed initial states, boosting real-task success by 10 to 34 points.
Backburner turns a plugged-in iPhone into a co-processor for a 24GB MacBook, boosting Qwen3-27B prefill up to 44% and extending context past 128k tokens.
Stanford and Google researchers introduce a diffusion method that spends compute only where detail matters, cutting token counts up to 4.6x.
Extend open-sourced Jevbox, a self-hosting document drive that uses hierarchical beam search instead of embeddings, with every answer tied to source permissions.
Microsoft Research shows that autonomous research agents get stuck in 'idea basins,' and a simple context-flush intervention lifts scores 66% over single runs.
A new vision backbone rewrites its own learning rules while scanning an image, hitting strong ImageNet, COCO, and ADE20K numbers with linear-time cost.
ByteShape ships ShapeLearn-quantized GGUFs of Qwen3.8-27B that fit on 12 GB GPUs and hit 176 tokens per second on an RTX 5090.
A ternary-quantized 27B vision-language model shrinks to 7.3 GB and specializes in writing ComfyUI image and video generation prompts.
A community post-trained Qwen variant uses RLOO to stop endless re-derivation, cutting 94K-token truncations and pushing LiveCodeBench from 83 to 90.
A Chinese Telecom lab released a 29B mixture-of-experts model with only 4B active parameters, 256K context, and strong agentic coding benchmarks.
A calibrated decision layer turns machine-learning interatomic potentials into probability engines that match full relaxation accuracy at a thirtieth of the compute.
A new open-source inference engine runs MiniMax H3 video generation locally on 8GB GPUs by streaming weights and adapting kernels to your hardware.
Stanford researchers introduce a compression-based framework that quantifies how close language models come to generating future research ideas, measured in bits.
Aleph Alpha drops a 78B mixture-of-experts reasoning model with 3.46B active parameters, 1M-token context, and Apache 2.0 weights, tuned for German-English sovereign deployments.
A new multi-agent framework uses a shared Git repository as working memory, letting parallel LLM workers inspect, extend, and build on each other's attempts.
A new synthesis framework builds 16,700 verifiable training environments across 40 professional domains, lifting a 35B Qwen model's GDPVal score from 45.5 to 73.6.
DSCODE wraps DeepSeek with a persistent shell, sub-agents that can hand off tasks, and an independent reviewer that approves changes from your instructions.
Red Hat AI shipped an FP8 build of NVIDIA's hybrid Mamba-MoE Lightning model, halving memory while keeping the 1M-token agent workhorse on a single GPU.