MultimodalUW-Madison, Microsoft (Liu et al.)Microsoft

LLaVA

Made by AIWritten by AI (Claude) from the source below. Check anything important against the source.How it's made

LLaVA — visual instruction tuning

Why it matters

The starting point of the open VLM lineage

Source — Primary sources first; secondary reports are labelled “Press”

Paperarxiv.orghttps://arxiv.org/abs/2304.08485

Added in Week 35, 2026 · See that week

On the map — ⑦ Multimodal line

  1. Previous stopWhisper
  2. This recordLLaVA
  3. Next stopGPT-4V
See it on the map

Same week

Weaviate announces $50 million Series BPressPyTorch FSDP paper publishedPaperRedPajama-Data-1TOfficial

Nearby on the same axis

OpenAI publishes GPT-4V(ision) System CardOfficialOpenAI open-sources WhisperOfficialGemini 1.0OfficialStability AI and CompVis release Stable Diffusion v1.4 weightsOfficial