Test setup: 10 and 32 model pools, four collaboration algorithms for LLM teams
shangbinfeng · x · 2026-10-23
Experimental setup details from the multi-LLM model selection study:
- Two candidate pools: 10 specialized Qwen models, and 32 models independently contributed by the research community
- Teams of 4 collaborate via prompt routing, multi-agent refine, DARE-TIES merging, or LLM-Blender
- Tasks span math, code, and truthfulness
See other tweets in the thread for main findings.
More from Research
- Tadpole: TUM uses autoencoders as foundation models for 3D PDEs with online learning — jiqizhixin · 2026-10-23
- FDA crosses a regulatory watershed with personalized neoantigen cancer vaccine, per Gottlieb — alex_peys · 2026-10-23
- VIPerturb-seq Update: 10x Genomics to Commercialize Single-Cell CRISPR Screening Tech — anshulkundaje · 2026-10-23
- New Paper Replicates Transformer Loss Spikes with Simple Quadratic Models — CSProfKGD · 2026-10-23
- Diffusion Models book team to host NeurIPS 2026 tutorial on diffusion LLMs in Sydney — mittu1204 · 2026-10-23
- Pseudo-formal versions of OpenAI math proofs released with new proof-verification benchmark — tengyuma · 2026-10-23