Fetching from the wire…
01
02
03
04
05
06
07
08
Top 5 · 2026-03-07 · source-backed
AI2's fully open 7B model replaces 75% of attention layers with Gated DeltaNet linear recurrence, matching OLMo 3 accuracy on MMLU with 49% fewer tokens. Nathan Lambert's analysis highlights the theoretical contribution: hybrid models formally solve problems that neither pure transformers nor linear RNNs can solve alone. Combined with Phi-4-reasoning-vision-15B (selective reasoning at 15B params, trained on 200B tokens in 4 days), small open models are producing genuinely competitive alternatives to API-dependent workflows. Architecture choice matters again, not just scale. (AI2 Blog | Interconnects)
Each link below shares sources, entities, or timing with this story.
Nathan Lambert's deep dive on OLMo Hybrid makes a compelling case that the transformer-only era may be ending. The paper formally proves hybrid models solve problems (related to code evaluation) that neither transformers nor GDN can solve alone. Combined with Qwen 3.5 and Kimi...
Fully open model using 3:1 DeltaNet-to-attention that matches Olmo 3 with 49% fewer tokens. Novel theoretical proof that hybrids are strictly more powerful than either transformers or RNNs alone. Trained on 512 GPUs. Full weights, checkpoints, and code released. AI2 Blog
1. Defend Against Vibeware DDoD — Behavioral process monitoring for AI-generated polyglot malware. Monitor Living Off Trusted Services patterns. Bitdefender 2. OS-Level Agent Sandbox — Kernel-level sandboxing (Seatbelt/Bubblewrap/AppContainer) for agentic workflows. Applicatio...
Three new components ship with it: GatedResidual, which mixes multiple residual streams with elementwise gating before each attention and MoE block; Qwen Sparse Attention, which scores compressed key blocks and selects contiguous token blocks while leaving the trailing block u...
1. Set Up Cursor Automations (intermediate) — Event-driven agents from PagerDuty/GitHub/Slack triggers with isolated sandboxes. Cursor Blog 2. Apply Context Engineering to Cut Agent Costs 60-80% (advanced) — Hierarchical token budgets, dynamic tool filtering (max 15), automati...
: [44] Reuters – Kalinowski [45] HN – Verification Debt [46] HN – LLM Tropes [47] HN – Qwen 3.5 Local [48] Interconnects – Olmo Hybrid ---
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.