Browse Comments — Raw (as collected)

Close reading of the corpus at each pipeline stage: raw → clean → relevant → coded.

↓ Export filtered CSV
Reading comments under one post — Amirhossein Hajianpour · AI Research & Models
DeepSeek AI beat OpenAI 15-25x on cost. We've created a new model type which beats Deepseek by the same margin. Earlier today, Modulate announced proven research into an AI architecture we've named t…
✕ clear post filter  ·  ← all posts
143 comments matched  ·  page 8 of 8
The ensemble approach is underexplored in the LLM space. Most of the industry defaulted to scaling monolithic models but the compute efficiency argument here is compelling. Routing distinct specialized models through an orchestrator rather than forcing one model to do everything mirrors how we think about mixture-of-experts at a higher abstraction level. Curious how ELM handles the coordination overhead and latency trade-offs across ensemble members.
AI/ML Engineer building RAG Pipelines ·… ⌕ thread
This would be tremendous for hotels wherein so much is already lost in STT
Advisor: Luxury Hotel Owners + Develope… ⌕ thread
This sounds interesting, will definitely try to dig more into it.
AI & Cloud Infrastructure engineer at O… ⌕ thread
← Prev 1 2 3 6 7 8 Next →